Video wird geladen...

Video konnte nicht geladen werden

Beim Laden dieses Videos ist ein Problem aufgetreten. Dies könnte an einem vorübergehenden Netzwerkproblem liegen oder das Video ist möglicherweise nicht verfügbar.

1.🚀 Baidu’s ERNIE 4.5 VL-28B-A3B-Thinking — an open-source MoE beast. With just 3B active params, it delivers SOTA multimodal reasoning: visual logic (MathVista 72.6), grounding, tool-calling, and more. The full family spans 12 models (0.3B → 424B). 🔗 Try it here: 💬 What’s your favorite “Thinking Mode” trick? #ERNIE45 #MultimodalAI

Helena AI

5,476 subscribers

53,547 Aufrufe • vor 8 Monaten •via X (Twitter)

Bildung Nachrichten & Politik Wissenschaft & Technologie

Anya Rossi• Live Now

Private livecam show

0 Kommentare

Keine Kommentare verfügbar

Kommentare vom Original-Post werden hier angezeigt

Ähnliche Videos

Six open-source LLMs. One sliding puzzle. A brutal test of long-horizon reasoning and tool calling. Five of them broke. One didn't. I gave each model a move_tile tool and a scrambled 3×3 board, then asked it to solve the puzzle through pure turn-by-turn reasoning. The deeper the scramble, the more brutal the search. Five runs per depth, best run kept. A model fails the round if it exceeds 6x the optimal move count. > Depth 5: Everyone solves it. Yawn. > Depth 10: GLM 5.1 melts down. 43 moves. Cut. > Depth 12: Gemma4 26B loses the plot, shuffling tiles in circles. Gone. > Depth 15: The wall. DeepSeek V4 Flash, out. DeepSeek V4 Pro, out. Gemma4, out again. GLM 5.1, out. Two survivors: Qwen3.6 35B-A3B, and Kimi K2.6 with an 11-move solve that looked like cheating. > Depth 18: Same two. Everyone else hallucinating tiles that weren't there. > Depth 22: Final boss. Kimi, flawless for five rounds, finally cracks. 81 moves. Still scrambled. DeepSeek V4 Pro limps home at 90. Qwen3.6 35B-A3B solves it in 36. The smallest model in the room. ~3B active params. Fits on a single 3090. It beat everything. Kimi was elegant. Qwen3.6 was unstoppable.

Six open-source LLMs. One sliding puzzle. A brutal test of long-horizon reasoning and tool calling. Five of them broke. One didn't. I gave each model a move_tile tool and a scrambled 3×3 board, then asked it to solve the puzzle through pure turn-by-turn reasoning. The deeper the scramble, the more brutal the search. Five runs per depth, best run kept. A model fails the round if it exceeds 6x the optimal move count. > Depth 5: Everyone solves it. Yawn. > Depth 10: GLM 5.1 melts down. 43 moves. Cut. > Depth 12: Gemma4 26B loses the plot, shuffling tiles in circles. Gone. > Depth 15: The wall. DeepSeek V4 Flash, out. DeepSeek V4 Pro, out. Gemma4, out again. GLM 5.1, out. Two survivors: Qwen3.6 35B-A3B, and Kimi K2.6 with an 11-move solve that looked like cheating. > Depth 18: Same two. Everyone else hallucinating tiles that weren't there. > Depth 22: Final boss. Kimi, flawless for five rounds, finally cracks. 81 moves. Still scrambled. DeepSeek V4 Pro limps home at 90. Qwen3.6 35B-A3B solves it in 36. The smallest model in the room. ~3B active params. Fits on a single 3090. It beat everything. Kimi was elegant. Qwen3.6 was unstoppable.

stevibe

16,434 Aufrufe • vor 3 Monaten

HERMES DESKTOP APP JUST GOT A NEW ULTRA MODE FOR DEEPSEEK V4 PRO – AND IT COSTS ABOUT $1 A DAY TO RUN If you're using Hermes desktop app, you're probably already running DeepSeek – it's one of the top models available. The update: → DeepSeek V4 Pro now has thinking mode, fast mode, and a brand new ultra mode → There's also an extra high mode, but the real upgrade is ultra The actual cost: roughly $1/day for a normal workflow. That's about $20-25/month – basically the same as before, just with a significantly faster mode unlocked. The verdict after a full day of testing: it flies through tool calling. Noticeably faster, same price range. If you're already paying for DeepSeek through Hermes, this is a free upgrade sitting in your settings right now. Update Hermes, grab the DeepSeek API if you don't have it yet, and try V4 Pro on ultra. Bookmark this, you'll want to come back to it later.

SCOTTY BEAM

24,032 Aufrufe • vor 1 Monat

NEWS: NVIDIA just announced Alpamayo, what CEO Jensen Huang calls the world’s first thinking, reasoning autonomous vehicle AI, launching on U.S. roads later this year, starting with the Mercedes CLA. Jensen: "It's trained end-to-end. Literally from camera in to actuation out; It reasons what action it is about to take, the reason by which is came about that action, and the trajectory." Alpamayo introduces Vision-Language-Action (VLA) models, which enable self-driving systems to interpret what they see, reason about complex driving scenarios, and generate driving actions. The platform includes large reasoning models, simulation tools for testing rare and edge-case scenarios, and open datasets for training and validation. NVIDIA says the approach improves transparency, safety, and robustness in autonomous systems, particularly in complex real-world environments, and supports progress toward higher levels of vehicle autonomy: "With a 10-billion-parameter architecture, Alpamayo 1 uses video input to generate trajectories alongside reasoning traces, showing the logic behind each decision. Developers can adapt Alpamayo 1 into smaller runtime models for vehicle development, or use it as a foundation for AV development tools such as reasoning-based evaluators and auto-labeling systems. Alpamayo 1 provides open model weights and open-source inferencing scripts. Future models in the family will feature larger parameter counts, more detailed reasoning capabilities, more input and output flexibility, and options for commercial usage."

NEWS: NVIDIA just announced Alpamayo, what CEO Jensen Huang calls the world’s first thinking, reasoning autonomous vehicle AI, launching on U.S. roads later this year, starting with the Mercedes CLA. Jensen: "It's trained end-to-end. Literally from camera in to actuation out; It reasons what action it is about to take, the reason by which is came about that action, and the trajectory." Alpamayo introduces Vision-Language-Action (VLA) models, which enable self-driving systems to interpret what they see, reason about complex driving scenarios, and generate driving actions. The platform includes large reasoning models, simulation tools for testing rare and edge-case scenarios, and open datasets for training and validation. NVIDIA says the approach improves transparency, safety, and robustness in autonomous systems, particularly in complex real-world environments, and supports progress toward higher levels of vehicle autonomy: "With a 10-billion-parameter architecture, Alpamayo 1 uses video input to generate trajectories alongside reasoning traces, showing the logic behind each decision. Developers can adapt Alpamayo 1 into smaller runtime models for vehicle development, or use it as a foundation for AV development tools such as reasoning-based evaluators and auto-labeling systems. Alpamayo 1 provides open model weights and open-source inferencing scripts. Future models in the family will feature larger parameter counts, more detailed reasoning capabilities, more input and output flexibility, and options for commercial usage."

Sawyer Merritt

1,604,362 Aufrufe • vor 7 Monaten

Nothing beats open-source! We now have a fully open-source coding agent that can compete with any proprietary solution out there: • 500+ available models • It doesn't train on your code • It doesn't charge a penny for tokens • It's open. You can change if you want. Kilo Code is the #1 app on OpenRouter right now, and the community is releasing new features on GitHub at rocket speed (about 20 releases in the last 30 days). I installed it in Visual Studio Code, but you can install it in pretty much any IDE. I need to spend more time with it, but here is a very interesting design choice: Most coding agents support two modes: an "agentic" mode for writing code and an "ask" mode for talking to the model without making any changes. Kilo Code supports 5 different modes: 1. Architect mode - Research, design, and planning 2. Debug mode - Problem solving and troubleshooting 3. Code mode - To write code. 4. Ask mode - To talk to the model. 5. Orchestrator mode - Combine every agent to solve problems. On paper, I really like this. I'll need more time to try out the specialized agents and see whether niching becomes an advantage. Here is the link to the website: Thanks to the Kilo for partnering with me on this post.

Nothing beats open-source! We now have a fully open-source coding agent that can compete with any proprietary solution out there: • 500+ available models • It doesn't train on your code • It doesn't charge a penny for tokens • It's open. You can change if you want. Kilo Code is the #1 app on OpenRouter right now, and the community is releasing new features on GitHub at rocket speed (about 20 releases in the last 30 days). I installed it in Visual Studio Code, but you can install it in pretty much any IDE. I need to spend more time with it, but here is a very interesting design choice: Most coding agents support two modes: an "agentic" mode for writing code and an "ask" mode for talking to the model without making any changes. Kilo Code supports 5 different modes: 1. Architect mode - Research, design, and planning 2. Debug mode - Problem solving and troubleshooting 3. Code mode - To write code. 4. Ask mode - To talk to the model. 5. Orchestrator mode - Combine every agent to solve problems. On paper, I really like this. I'll need more time to try out the specialized agents and see whether niching becomes an advantage. Here is the link to the website: Thanks to the Kilo for partnering with me on this post.

Santiago

73,062 Aufrufe • vor 9 Monaten

Today, we introduce HunyuanImage 3.0-Instruct, a native multimodal model focusing on image-editing by integrating visual understanding with precise image synthesis! 🚀 It understands input images and reasons before generating images. Built on an 80B-parameter MoE architecture (13B activated), it natively unifies deep multimodal comprehension and high-fidelity generation. 🧠 A "Thinking" Model with Native CoT & MixGRPO: The model doesn’t just execute commands, it processes them through a Native Chain-of-Thought (CoT) schema. Enhanced by our self-developed MixGRPO algorithm, it reasons through complex instructions to achieve flawless intent alignment and human-preference consistency. 🎨 Precise Editing & Multi-Image Fusion: The model enables accurate image editing by adding, removing, or modifying elements while keeping non-target areas perfectly intact. It also excels at seamless multi-image fusion, synthesizing complex scenes by extracting and blending elements from multiple sources into a unified, consistent output. 🏆 SOTA Performance: HunyuanImage 3.0-Instruct sets a new benchmark in visual quality and alignment, delivering performance that matches leading proprietary models. We aim to enable the community to explore new ideas with a state-of-the-art foundation model, fostering a dynamic and vibrant image generation ecosystem. 🛠️🎨 💻Try it at (PC only):

Today, we introduce HunyuanImage 3.0-Instruct, a native multimodal model focusing on image-editing by integrating visual understanding with precise image synthesis! 🚀 It understands input images and reasons before generating images. Built on an 80B-parameter MoE architecture (13B activated), it natively unifies deep multimodal comprehension and high-fidelity generation. 🧠 A "Thinking" Model with Native CoT & MixGRPO: The model doesn’t just execute commands, it processes them through a Native Chain-of-Thought (CoT) schema. Enhanced by our self-developed MixGRPO algorithm, it reasons through complex instructions to achieve flawless intent alignment and human-preference consistency. 🎨 Precise Editing & Multi-Image Fusion: The model enables accurate image editing by adding, removing, or modifying elements while keeping non-target areas perfectly intact. It also excels at seamless multi-image fusion, synthesizing complex scenes by extracting and blending elements from multiple sources into a unified, consistent output. 🏆 SOTA Performance: HunyuanImage 3.0-Instruct sets a new benchmark in visual quality and alignment, delivering performance that matches leading proprietary models. We aim to enable the community to explore new ideas with a state-of-the-art foundation model, fostering a dynamic and vibrant image generation ecosystem. 🛠️🎨 💻Try it at (PC only):

Tencent Hy

126,634 Aufrufe • vor 7 Monaten

Alibaba just dropped Qwen3.5-397B-A17B and there's a lot to unpack. 397B params, 17B active per forward pass. Sparse MoE done right. But the real story isn't the size—it's the architecture choices. The MoE Design Most MoE models feel like bolt-ons. Qwen 3.5's sparse activation is native—only 4.3% of parameters fire per token. That's how you get trillion-parameter-class performance without trillion-parameter inference costs. The 0.8 RMB/million tokens pricing isn't subsidized; it's structurally earned. Native Multimodal, Not Glued-On This is a vision-language model from the ground up. Heterogeneous architecture—separate processing pipelines for text, image, video that fuse early. Not a vision encoder slapped onto an LLM. The result: 90.8 on OmniDocBench, 79.0 on MMMU-Pro. Document understanding and visual reasoning without the usual brittleness. The Context Window Reality Qwen3.5-Plus (the hosted version) ships with 1M tokens by default. That's not a marketing number—they're actually positioning it for long-document workflows. With built-in adaptive tool use, it's clearly aimed at agentic automation, not just chat. What Actually Impressed Me • FP8 native pipeline: ~50% activation memory reduction • Async RL framework for continuous refinement—training and inference workloads separated • 201 languages (up from 119), 250k vocab for better low-resource encoding • Apache 2.0 license. Full weights on HuggingFace and ModelScope. The Benchmark Context 76.4 on SWE-bench Verified puts it in the range where it can handle real debugging workflows. 72.9 on BFCL v4 for agentic tool use. 88.4 on GPQA Diamond. These aren't SOTA in isolation, but the breadth is unusual—strong across reasoning, coding, multimodal, and agentic tasks. The Honest Caveat I haven't stress-tested the 1M context for needle-in-haystack retrieval yet. And "native multimodal" claims need real-world torture testing—PDFs with tables, charts, mixed layouts. Benchmarks are benchmarks. Bottom Line This isn't just another model release. It's a bet on efficient scale: big model capabilities, small active compute, open weights. At 1/18th the cost of Gemini 3 Pro, it's going to force pricing conversations across the board.

Alibaba just dropped Qwen3.5-397B-A17B and there's a lot to unpack. 397B params, 17B active per forward pass. Sparse MoE done right. But the real story isn't the size—it's the architecture choices. The MoE Design Most MoE models feel like bolt-ons. Qwen 3.5's sparse activation is native—only 4.3% of parameters fire per token. That's how you get trillion-parameter-class performance without trillion-parameter inference costs. The 0.8 RMB/million tokens pricing isn't subsidized; it's structurally earned. Native Multimodal, Not Glued-On This is a vision-language model from the ground up. Heterogeneous architecture—separate processing pipelines for text, image, video that fuse early. Not a vision encoder slapped onto an LLM. The result: 90.8 on OmniDocBench, 79.0 on MMMU-Pro. Document understanding and visual reasoning without the usual brittleness. The Context Window Reality Qwen3.5-Plus (the hosted version) ships with 1M tokens by default. That's not a marketing number—they're actually positioning it for long-document workflows. With built-in adaptive tool use, it's clearly aimed at agentic automation, not just chat. What Actually Impressed Me • FP8 native pipeline: ~50% activation memory reduction • Async RL framework for continuous refinement—training and inference workloads separated • 201 languages (up from 119), 250k vocab for better low-resource encoding • Apache 2.0 license. Full weights on HuggingFace and ModelScope. The Benchmark Context 76.4 on SWE-bench Verified puts it in the range where it can handle real debugging workflows. 72.9 on BFCL v4 for agentic tool use. 88.4 on GPQA Diamond. These aren't SOTA in isolation, but the breadth is unusual—strong across reasoning, coding, multimodal, and agentic tasks. The Honest Caveat I haven't stress-tested the 1M context for needle-in-haystack retrieval yet. And "native multimodal" claims need real-world torture testing—PDFs with tables, charts, mixed layouts. Benchmarks are benchmarks. Bottom Line This isn't just another model release. It's a bet on efficient scale: big model capabilities, small active compute, open weights. At 1/18th the cost of Gemini 3 Pro, it's going to force pricing conversations across the board.

Bo Wang

13,221 Aufrufe • vor 6 Monaten

Introducing "Building with Llama 4." This short course is created with Meta AI at Meta, and taught by Amit Sangani, Director of Partner Engineering for Meta’s AI team. Meta’s new Llama 4 has added three new models and introduced the Mixture-of-Experts (MoE) architecture to its family of open-weight models, making them more efficient to serve. In this course, you’ll work with two of the three new models introduced in Llama 4. First is Maverick, a 400B parameter model, with 128 experts and 17B active parameters. Second is Scout, a 109B parameter model with 16 experts and 17B active parameters. Maverick and Scout support long context windows of up to a million tokens and 10M tokens, respectively. The latter is enough to support directly inputting even fairly large GitHub repos for analysis! In hands-on lessons, you’ll build apps using Llama 4’s new multimodal capabilities including reasoning across multiple images and image grounding, in which you can identify elements in images. You’ll also use the official Llama API, work with Llama 4’s long-context abilities, and learn about Llama’s newest open-source tools: its prompt optimization tool that automatically improves system prompts and synthetic data kit that generates high-quality datasets for fine-tuning. If you need an open model, Llama is a great option, and the Llama 4 family is an important part of any GenAI developer's toolkit. Through this course, you’ll learn to call Llama 4 via API, use its optimization tools, and build features that span text, images, and large context. Please sign up here:

Introducing "Building with Llama 4." This short course is created with Meta AI at Meta, and taught by Amit Sangani, Director of Partner Engineering for Meta’s AI team. Meta’s new Llama 4 has added three new models and introduced the Mixture-of-Experts (MoE) architecture to its family of open-weight models, making them more efficient to serve. In this course, you’ll work with two of the three new models introduced in Llama 4. First is Maverick, a 400B parameter model, with 128 experts and 17B active parameters. Second is Scout, a 109B parameter model with 16 experts and 17B active parameters. Maverick and Scout support long context windows of up to a million tokens and 10M tokens, respectively. The latter is enough to support directly inputting even fairly large GitHub repos for analysis! In hands-on lessons, you’ll build apps using Llama 4’s new multimodal capabilities including reasoning across multiple images and image grounding, in which you can identify elements in images. You’ll also use the official Llama API, work with Llama 4’s long-context abilities, and learn about Llama’s newest open-source tools: its prompt optimization tool that automatically improves system prompts and synthetic data kit that generates high-quality datasets for fine-tuning. If you need an open model, Llama is a great option, and the Llama 4 family is an important part of any GenAI developer's toolkit. Through this course, you’ll learn to call Llama 4 via API, use its optimization tools, and build features that span text, images, and large context. Please sign up here:

Andrew Ng

67,846 Aufrufe • vor 1 Jahr

Love OpenClaw but hate the token burn? 💸 Running a 24/7 agent on GPT-4/Claude is overkill. You don't need SOTA reasoning to handle a greeting or a simple lookup. LLMRouter 🩷 OpenClaw The first production-ready, agentic router designed to plug directly into OpenClaw. LLMRouter fully supports Multimodal, Memory-Equipped routing that adapts 100% to your needs—compatible with FREE open-source models. The Logic is Simple:🔹 Simple query → Cheap/Local model 🔹 Complex reasoning → SOTA model (GPT-4/Claude 3.5) 🔹 Multimodal input → Vision/Audio specialized model Why this isn't just a switch: 📉 30–50% drop in inference costs 🧠 Zero loss in response quality 🔓 100% compatible with OpenAI-style APIs 🚀 Deploy in Seconds General Usage: Get the library and serve any model: pip install llmrouter-lib llmrouter serve OpenClaw Native Integration: Want the full agent experience? LLMRouter built a dedicated integration for OpenClaw users: LLMRouter Resources: 🔗 Repo: 📦 PyPI: 🤝 Works with: Route smarter. Train your own. Pay less. More on LLMRouter: Most routers are static if/else. LLMRouter is an intelligent, learning system. 🤖 Agentic & Memory-Aware: Decisions aren't stateless. We use RAG-powered memory to route based on context and history. 👤 Fully Personalized: It learns from your usage patterns via RL feedback loops. 🔬 Research-Grade: Switch between 16+ routing strategies (KNN, SVM, BERT, Graph, RL) with a single flag.

Love OpenClaw but hate the token burn? 💸 Running a 24/7 agent on GPT-4/Claude is overkill. You don't need SOTA reasoning to handle a greeting or a simple lookup. LLMRouter 🩷 OpenClaw The first production-ready, agentic router designed to plug directly into OpenClaw. LLMRouter fully supports Multimodal, Memory-Equipped routing that adapts 100% to your needs—compatible with FREE open-source models. The Logic is Simple:🔹 Simple query → Cheap/Local model 🔹 Complex reasoning → SOTA model (GPT-4/Claude 3.5) 🔹 Multimodal input → Vision/Audio specialized model Why this isn't just a switch: 📉 30–50% drop in inference costs 🧠 Zero loss in response quality 🔓 100% compatible with OpenAI-style APIs 🚀 Deploy in Seconds General Usage: Get the library and serve any model: pip install llmrouter-lib llmrouter serve OpenClaw Native Integration: Want the full agent experience? LLMRouter built a dedicated integration for OpenClaw users: LLMRouter Resources: 🔗 Repo: 📦 PyPI: 🤝 Works with: Route smarter. Train your own. Pay less. More on LLMRouter: Most routers are static if/else. LLMRouter is an intelligent, learning system. 🤖 Agentic & Memory-Aware: Decisions aren't stateless. We use RAG-powered memory to route based on context and history. 👤 Fully Personalized: It learns from your usage patterns via RL feedback loops. 🔬 Research-Grade: Switch between 16+ routing strategies (KNN, SVM, BERT, Graph, RL) with a single flag.

Jiaxuan You

31,451 Aufrufe • vor 6 Monaten

The Kimi AI Model And What If This Carnegie Mellon, PhD student was incentivized to open source AI in the US? The answer is Kimi would have been an open source US model. Yang Zhilin Moonshot’s founder and maker of the Kimi series turned his deep research expertise into one of the world’s most capable open AI systems. After earning his PhD at Carnegie Mellon under leading researchers and interning at Google Brain and Meta, he returned to China and co-founded Moonshot AI in March 2023 with Tsinghua classmates Zhou Xinyu and Wu Yuxin. Why? The US VC world and tax structure did not favor Zhilin’s proposal to open source the AI models as a strategy. So he left. He named the company after his favorite Pink Floyd album, reflecting his ambitious vision for scalable. Yang assembled a core technical team of inventors behind breakthroughs like Transformer-XL and RoPE. Together they focused on turning massive compute into efficient intelligence through innovative architectures. The journey began with the Kimi chatbot in October 2023, rapidly scaling context from 200,000 to millions of characters. This evolved into the Kimi series: K1.5 matched top reasoning models, K2 introduced a 1-trillion-parameter Mixture-of-Experts design trained on 15.5 trillion tokens and released openly, and K2 Thinking added advanced agentic capabilities. Kimi K3 represents the pinnacle, this 2.8-trillion-parameter model uses a sparse MoE architecture with 896 experts (only 16 active per token), new Kimi Delta Attention and attention residuals for efficiency, and a 1-million-token context window. It delivers frontier performance in long-horizon coding, reasoning, and multimodal tasks at competitive cost, with weights set for open release. Yang’s approach emphasizes openness, efficiency, and continuous self-improvement — enabling solo developers and teams to achieve what once required massive resources. By sharing technical insights in public talks, he has accelerated global progress toward more accessible, powerful AI. Imagine if we held open source higher than the fear theater games of Anthropic? We are chasing out some of the best minds. This is how you lose…

The Kimi AI Model And What If This Carnegie Mellon, PhD student was incentivized to open source AI in the US? The answer is Kimi would have been an open source US model. Yang Zhilin Moonshot’s founder and maker of the Kimi series turned his deep research expertise into one of the world’s most capable open AI systems. After earning his PhD at Carnegie Mellon under leading researchers and interning at Google Brain and Meta, he returned to China and co-founded Moonshot AI in March 2023 with Tsinghua classmates Zhou Xinyu and Wu Yuxin. Why? The US VC world and tax structure did not favor Zhilin’s proposal to open source the AI models as a strategy. So he left. He named the company after his favorite Pink Floyd album, reflecting his ambitious vision for scalable. Yang assembled a core technical team of inventors behind breakthroughs like Transformer-XL and RoPE. Together they focused on turning massive compute into efficient intelligence through innovative architectures. The journey began with the Kimi chatbot in October 2023, rapidly scaling context from 200,000 to millions of characters. This evolved into the Kimi series: K1.5 matched top reasoning models, K2 introduced a 1-trillion-parameter Mixture-of-Experts design trained on 15.5 trillion tokens and released openly, and K2 Thinking added advanced agentic capabilities. Kimi K3 represents the pinnacle, this 2.8-trillion-parameter model uses a sparse MoE architecture with 896 experts (only 16 active per token), new Kimi Delta Attention and attention residuals for efficiency, and a 1-million-token context window. It delivers frontier performance in long-horizon coding, reasoning, and multimodal tasks at competitive cost, with weights set for open release. Yang’s approach emphasizes openness, efficiency, and continuous self-improvement — enabling solo developers and teams to achieve what once required massive resources. By sharing technical insights in public talks, he has accelerated global progress toward more accessible, powerful AI. Imagine if we held open source higher than the fear theater games of Anthropic? We are chasing out some of the best minds. This is how you lose…

Brian Roemmele

37,185 Aufrufe • vor 1 Monat

Introducing PhoneLLM, an open model for voice agents. GPT 5.6 Terra performance on typical voice agent tasks at 1/3 the latency and 1/18 the cost. For voice agents, we need models that are both very low latency and very good at tool calling and instruction following. There's a trade-off here, and we often have to compromise on either latency or capability when building voice agents. With PhoneLLM (and the training and data stack that made this model possible) we're fixing this problem. For the last couple of years, most of the effort in frontier model development has gone towards leveraging test-time compute. Which is awesome! Models of all shapes and sizes are available that perform really, really well ... if you have "thinking" turned on for your model. But if you need your agent to respond at voice conversation speed, you can't use thinking models. PhoneLLM is a full-weights fine-tune of NVIDIA Nemotron Nano 30B. We trained on a wide range of real-world telephone and customer support use cases. The training focused on taking the excellent Nano 30B base capabilities and teaching the model to do typical voice agent tasks with thinking disabled. The results are really good: accurate tool calling and concise, on-topic responses in long conversations. And fast: TTFAT measured server-side is <100ms if you run PhoneLLM on a lightly loaded B200. :-) But seriously, when we characterize model latency, we do it with full, end-to-end, batched request simulations using real Pipecat voice agent pipelines. You can serve more than 80 concurrent agents on a single B200 with P95 end-to-end TTFAT <600ms. Including network overhead. That's an LLM cost-per-minute around $0.0025. (1/4 of a cent.) At a latency lower than any third-party API offers today. More details about this model, including weights on Hugging Face, how to spin it up with one click on Modal, and a starter project repo you can clone, are in the thread ...

Introducing PhoneLLM, an open model for voice agents. GPT 5.6 Terra performance on typical voice agent tasks at 1/3 the latency and 1/18 the cost. For voice agents, we need models that are both very low latency and very good at tool calling and instruction following. There's a trade-off here, and we often have to compromise on either latency or capability when building voice agents. With PhoneLLM (and the training and data stack that made this model possible) we're fixing this problem. For the last couple of years, most of the effort in frontier model development has gone towards leveraging test-time compute. Which is awesome! Models of all shapes and sizes are available that perform really, really well ... if you have "thinking" turned on for your model. But if you need your agent to respond at voice conversation speed, you can't use thinking models. PhoneLLM is a full-weights fine-tune of NVIDIA Nemotron Nano 30B. We trained on a wide range of real-world telephone and customer support use cases. The training focused on taking the excellent Nano 30B base capabilities and teaching the model to do typical voice agent tasks with thinking disabled. The results are really good: accurate tool calling and concise, on-topic responses in long conversations. And fast: TTFAT measured server-side is <100ms if you run PhoneLLM on a lightly loaded B200. :-) But seriously, when we characterize model latency, we do it with full, end-to-end, batched request simulations using real Pipecat voice agent pipelines. You can serve more than 80 concurrent agents on a single B200 with P95 end-to-end TTFAT <600ms. Including network overhead. That's an LLM cost-per-minute around $0.0025. (1/4 of a cent.) At a latency lower than any third-party API offers today. More details about this model, including weights on Hugging Face, how to spin it up with one click on Modal, and a starter project repo you can clone, are in the thread ...

kwindla

84,404 Aufrufe • vor 10 Stunden

If you use LLM-as-judge, this one is for you. (bookmark it) Most teams validate their agent's outputs by calling a frontier model as the judge. It works, until it doesn't. Three problems stack up fast: → Cost: you're hitting a frontier API on every turn, every tool call, every response. In production that burns millions. → Latency: bigger models, remote calls, slow reasoning on every check. → Blind spots: frontier models don't actually know your domain. In finance, insurance, or healthcare, they miss the keywords and principles your work depends on. So I walk through a different approach: train your own small LLM judge. Instead of a giant model, you start with a small one and let the system generate the training data for you. It decomposes your domain, samples synthetic examples, runs them through a debate arena where judges reach consensus, then trains on the refined set. The result is a judge that's cheaper, faster, and more accurate on your data than Gemini, Claude, or GPT, with an OpenAI-compatible endpoint you can even deploy on-prem. I show the whole thing end to end, using a Claude Code plugin and a web interface, with a real insurance RAG grounding evaluator as the example. You can get the plugin here: Here's the full breakdown: 00:00 - Intro 00:12 - Three problems with using frontier LLMs as judges 01:05 - A different approach: train your own small judge 01:31 - How it works (synthetic data and a debate arena) 02:50 - Installing the Claude Code plugin 04:03 - Defining your task with /eval 04:34 - Example: an insurance RAG grounding evaluator 05:51 - Kicking it off and giving early feedback 06:26 - Choosing labels, domain, and strictness 08:30 - The web interface and dashboard 09:52 - Bringing your own example data (optional) 10:26 - The finished model: endpoint, accuracy, and speed 11:16 - Control, on-prem deployment, and interpretability 11:57 - Benchmarks vs frontier models and the GitHub repo 12:30 - Outro I worked with the Plurai team on this. Thanks for sponsoring the video.

If you use LLM-as-judge, this one is for you. (bookmark it) Most teams validate their agent's outputs by calling a frontier model as the judge. It works, until it doesn't. Three problems stack up fast: → Cost: you're hitting a frontier API on every turn, every tool call, every response. In production that burns millions. → Latency: bigger models, remote calls, slow reasoning on every check. → Blind spots: frontier models don't actually know your domain. In finance, insurance, or healthcare, they miss the keywords and principles your work depends on. So I walk through a different approach: train your own small LLM judge. Instead of a giant model, you start with a small one and let the system generate the training data for you. It decomposes your domain, samples synthetic examples, runs them through a debate arena where judges reach consensus, then trains on the refined set. The result is a judge that's cheaper, faster, and more accurate on your data than Gemini, Claude, or GPT, with an OpenAI-compatible endpoint you can even deploy on-prem. I show the whole thing end to end, using a Claude Code plugin and a web interface, with a real insurance RAG grounding evaluator as the example. You can get the plugin here: Here's the full breakdown: 00:00 - Intro 00:12 - Three problems with using frontier LLMs as judges 01:05 - A different approach: train your own small judge 01:31 - How it works (synthetic data and a debate arena) 02:50 - Installing the Claude Code plugin 04:03 - Defining your task with /eval 04:34 - Example: an insurance RAG grounding evaluator 05:51 - Kicking it off and giving early feedback 06:26 - Choosing labels, domain, and strictness 08:30 - The web interface and dashboard 09:52 - Bringing your own example data (optional) 10:26 - The finished model: endpoint, accuracy, and speed 11:16 - Control, on-prem deployment, and interpretability 11:57 - Benchmarks vs frontier models and the GitHub repo 12:30 - Outro I worked with the Plurai team on this. Thanks for sponsoring the video.

Akshay 🚀

30,430 Aufrufe • vor 1 Monat

We’re excited to announce the release and open-source of HunyuanImage 3.0 — the largest and most powerful open-source text-to-image model to date, with over 80 billion total parameters, of which 13 billion are activated per token during inference.The effect is completely comparable to the industry’s flagship closed-source model.🚀🚀🚀 HunyuanImage 3.0 originates from our internally developed native multimodal large language model, with fine-tuning and post-training focused on text-to-image generation. This unique foundation gives the model a powerful set of capabilities: ✅Reason with world knowledge ✅Understand complex, thousand-word prompts ✅Generate precise text within images Different from traditional DiT architecture image generation models, HunyuanImage 3.0’s MoE architecture uses a Transfusion-based approach to deeply couple Diffusion and LLM training for a single, powerful system. Built on Hunyuan-A13B, HunyuanImage 3.0 was trained on a massive dataset: 5 billion image-text pairs, video frames, interleaved image-text data, and 6 trillion tokens of text corpora. This hybrid training across multimodal generation, understanding, and LLM capabilities allows the model to seamlessly integrate multiple tasks. Whether you're an illustrator, designer, or creator, this is built to slash your workflow from hours to minutes. HunyuanImage 3.0 can generate intricate text, detailed comics, expressive emojis, and lively, engaging illustrations for educational content. The current release focuses solely on text-to-image generation and future updates will include image-to-image, image editing, multi-turn interaction, and more. 👉🏻Try it now: 🔗GitHub: 🤗Hugging Face:

We’re excited to announce the release and open-source of HunyuanImage 3.0 — the largest and most powerful open-source text-to-image model to date, with over 80 billion total parameters, of which 13 billion are activated per token during inference.The effect is completely comparable to the industry’s flagship closed-source model.🚀🚀🚀 HunyuanImage 3.0 originates from our internally developed native multimodal large language model, with fine-tuning and post-training focused on text-to-image generation. This unique foundation gives the model a powerful set of capabilities: ✅Reason with world knowledge ✅Understand complex, thousand-word prompts ✅Generate precise text within images Different from traditional DiT architecture image generation models, HunyuanImage 3.0’s MoE architecture uses a Transfusion-based approach to deeply couple Diffusion and LLM training for a single, powerful system. Built on Hunyuan-A13B, HunyuanImage 3.0 was trained on a massive dataset: 5 billion image-text pairs, video frames, interleaved image-text data, and 6 trillion tokens of text corpora. This hybrid training across multimodal generation, understanding, and LLM capabilities allows the model to seamlessly integrate multiple tasks. Whether you're an illustrator, designer, or creator, this is built to slash your workflow from hours to minutes. HunyuanImage 3.0 can generate intricate text, detailed comics, expressive emojis, and lively, engaging illustrations for educational content. The current release focuses solely on text-to-image generation and future updates will include image-to-image, image editing, multi-turn interaction, and more. 👉🏻Try it now: 🔗GitHub: 🤗Hugging Face:

Tencent Hy

412,880 Aufrufe • vor 11 Monaten

We've officially released and open-sourced HunyuanImage 2.1, our latest text-to-image model. The new model delivers on our commitment to balancing performance and quality. With native 2K image generation, HunyuanImage 2.1 is an advanced open-source text-to-image model.🎨 ✨ New in 2.1: 🔹Advanced Semantics: Supports ultra-long and complex prompts of up to 1000 tokens, and precisely controls the generation of multiple subjects in a single image. 🔹Precise Chinese and English Text Rendering with seamless image–text integration: The model naturally integrates text into images, making it suitable for a wide range of applications such as product covers, illustrations, and poster design to meet the needs of various fields. 🔹Rich Styles and High Aesthetic: Capable of generating images in various styles—including photorealistic portraits, comics, and vinyl figures—it delivers outstanding visual appeal and artistic quality. 🔹High-Quality Generation: Efficiently produces ultra-high-definition (2K) images in the same time other models take to generate a 1K image. HunyuanImage 2.1 uses two text encoders: a multimodal large language model (MLLM) to improve the model's image and text alignment capabilities, and a multi-language character-aware encoder to improve text rendering capabilities. The model is a single- and double-stream diffusion transformer with 17B parameters. We've also open-sourced the weights of the the accelerated version with meanflow which reduces inference steps from 100 to just 8, and PromptEnhancer, the first industrial-grade rewriting model that enhances your prompts for more nuanced and expressive image generation. Now, creators turn complex ideas—like posters with slogans or multi-panel comics—into visuals faster than ever. We’re just getting started. Stay tuned for our native multimodal image generation model coming soon. 🌐Website: 🔗Github: 🤗Hugging Face: ✨Hugging Face Demo:

We've officially released and open-sourced HunyuanImage 2.1, our latest text-to-image model. The new model delivers on our commitment to balancing performance and quality. With native 2K image generation, HunyuanImage 2.1 is an advanced open-source text-to-image model.🎨 ✨ New in 2.1: 🔹Advanced Semantics: Supports ultra-long and complex prompts of up to 1000 tokens, and precisely controls the generation of multiple subjects in a single image. 🔹Precise Chinese and English Text Rendering with seamless image–text integration: The model naturally integrates text into images, making it suitable for a wide range of applications such as product covers, illustrations, and poster design to meet the needs of various fields. 🔹Rich Styles and High Aesthetic: Capable of generating images in various styles—including photorealistic portraits, comics, and vinyl figures—it delivers outstanding visual appeal and artistic quality. 🔹High-Quality Generation: Efficiently produces ultra-high-definition (2K) images in the same time other models take to generate a 1K image. HunyuanImage 2.1 uses two text encoders: a multimodal large language model (MLLM) to improve the model's image and text alignment capabilities, and a multi-language character-aware encoder to improve text rendering capabilities. The model is a single- and double-stream diffusion transformer with 17B parameters. We've also open-sourced the weights of the the accelerated version with meanflow which reduces inference steps from 100 to just 8, and PromptEnhancer, the first industrial-grade rewriting model that enhances your prompts for more nuanced and expressive image generation. Now, creators turn complex ideas—like posters with slogans or multi-panel comics—into visuals faster than ever. We’re just getting started. Stay tuned for our native multimodal image generation model coming soon. 🌐Website: 🔗Github: 🤗Hugging Face: ✨Hugging Face Demo:

Tencent Hy

89,257 Aufrufe • vor 11 Monaten

China just released an open source AI model that matches the best closed models from OpenAI and Anthropic. Gavin Baker explained exactly how they did it and the answer should concern every American AI lab. The model is called GLM 5.2. It was built by Z. AI. You get 744 billion parameters, 1 million token context window and its MIT license, meaning anyone can download it, fork it, build a company on it, with no restrictions and no Dario. It scored 51 points on the artificial analysis intelligence index. The highest score any open weight model has ever achieved. It beat GPT 5.5 on the frontier software engineering benchmark. It trails Claude Opus 4.8 by less than one percentage point. And it costs 85% less to run than GPT 5.5 for comparable performance. Gavin Baker said on the All-In podcast that this model has challenged some of his beliefs. Then he explained how China built it. The method is called distillation. Just think of tens of thousands of phones and computers running simultaneously, all hitting the frontier model APIs through masked accounts, asking specific questions, and harvesting what happens inside the model when it answers. Every reasoning step, every token. The entire thinking process gets recorded and fed back into the Chinese model during training. It is a cheat sheet. It is the answer key to the exam. And here is the part that should worry everyone. Sacks said it plainly. China was already nine months behind American models. But now that GLM 5.2 is good enough to run its own reinforcement learning, it can improve itself without needing to distill from American models anymore. The cheat sheet let them get close enough to start writing their own answers. Sacks said we are six months behind on the model and 24 months behind on silicon and they are only a few months behind in total. The Z. AI founder told Elon Musk directly that open weight fable-level capability will be here before Q1 2027. Every restriction Anthropic lobbied for, every self-imposed safety guardrail, every month of delay in releasing American frontier models accelerated this. The Chinese labs were not under those restrictions. They were not going to wait. The composable model future Gavin described, where every enterprise runs a frontier model alongside their own fine-tuned open weight model, is coming regardless of what American labs do next. The question is just whether the open weight half of that stack is American or Chinese. Right now it is Chinese. WATCH THE FULL PODCAST ON The All-In Podcast

China just released an open source AI model that matches the best closed models from OpenAI and Anthropic. Gavin Baker explained exactly how they did it and the answer should concern every American AI lab. The model is called GLM 5.2. It was built by Z. AI. You get 744 billion parameters, 1 million token context window and its MIT license, meaning anyone can download it, fork it, build a company on it, with no restrictions and no Dario. It scored 51 points on the artificial analysis intelligence index. The highest score any open weight model has ever achieved. It beat GPT 5.5 on the frontier software engineering benchmark. It trails Claude Opus 4.8 by less than one percentage point. And it costs 85% less to run than GPT 5.5 for comparable performance. Gavin Baker said on the All-In podcast that this model has challenged some of his beliefs. Then he explained how China built it. The method is called distillation. Just think of tens of thousands of phones and computers running simultaneously, all hitting the frontier model APIs through masked accounts, asking specific questions, and harvesting what happens inside the model when it answers. Every reasoning step, every token. The entire thinking process gets recorded and fed back into the Chinese model during training. It is a cheat sheet. It is the answer key to the exam. And here is the part that should worry everyone. Sacks said it plainly. China was already nine months behind American models. But now that GLM 5.2 is good enough to run its own reinforcement learning, it can improve itself without needing to distill from American models anymore. The cheat sheet let them get close enough to start writing their own answers. Sacks said we are six months behind on the model and 24 months behind on silicon and they are only a few months behind in total. The Z. AI founder told Elon Musk directly that open weight fable-level capability will be here before Q1 2027. Every restriction Anthropic lobbied for, every self-imposed safety guardrail, every month of delay in releasing American frontier models accelerated this. The Chinese labs were not under those restrictions. They were not going to wait. The composable model future Gavin described, where every enterprise runs a frontier model alongside their own fine-tuned open weight model, is coming regardless of what American labs do next. The question is just whether the open weight half of that stack is American or Chinese. Right now it is Chinese. WATCH THE FULL PODCAST ON The All-In Podcast

Ihtesham Ali

86,452 Aufrufe • vor 2 Monaten

HERMES AGENT SUPPORTS 300+ MODELS. PICKING THE RIGHT ONE PER TASK IS THE DIFFERENCE BETWEEN $5/MONTH AND $50. STARTING OUT: Claude Sonnet 4.6. official recommendation from Nous Research. "the model this project was built and tested with." strong reasoning. reliable tool calling. mid-range pricing. PREMIUM TIER: Claude Opus 4.8. best coding benchmarks available. self-correcting reasoning. catches its own mistakes. 1M context. use for demanding tasks where quality matters. GPT-5.5. #1 Chatbot Arena. #1 GPQA Diamond reasoning (94.1%). #1 creative writing. 2M context. handles entire codebases in one pass. Grok 4.30. the only frontier model with live X firehose access. real-time social data, breaking news, market sentiment. connects via Grok OAuth. no separate API key. Grok-Composer-2.5-Fast (v0.17.0). Cursor's coding model. 200K context. available through your Grok subscription via OAuth. no extra cost if you already pay for Grok. MID-RANGE TIER: Claude Sonnet 4.6. best balance of quality and cost for daily use. strongest prose and tool calling in this tier. Gemini 2.5 Pro. Google Search grounding built in. cites sources. verifies claims. pulls current data. 2M context. best for research-heavy workflows. GPT-4.1. reliable tool calling. solid general reasoning. good middle ground when you need OpenAI compatibility. BUDGET TIER: Claude Haiku 4.5. fastest Anthropic model. cheapest paid Claude option. strong at classification, routing, simple queries. use for auxiliary tasks: compression, vision, web extraction, approval scoring. DeepSeek V4. best cost-to-quality ratio in the market. 90% cache discount on repeated context. use for sub-agents and bulk parallel work. DeepSeek V4 Flash. cheapest paid model worth using. 1M context. MIT license. self-hostable. use for cron jobs, monitoring, routine searches. MiniMax M3. Nous Research and MiniMax collaborating on optimization. 1M context via lightning attention. 59% SWE-Bench Pro. beats several premium models on coding. one of the most-used models inside Hermes. FREE / LOCAL: Qwen 3.5 27B via Ollama. 16GB VRAM. reliable tool calling. best free local model for Hermes as of mid-2026. Qwen 3 8B. 8GB VRAM. fits a $7 VPS. handles routine tasks at zero API cost. Llama 4 Maverick. best open-weight tool calling. 1M context. needs more VRAM but strongest local option. HOW TO ASSIGN MODELS: main model: Desktop app / Dashboard → Models → switch sub-agent model: set in Desktop app, Dashboard, or config.yaml: delegation: model: "deepseek/deepseek-v4" auxiliary models (compression, vision, web extract): Desktop app / Dashboard → Models → Auxiliary Haiku 4.5 or Gemini Flash work well here. saves significantly when your main model is premium. per-profile: each Hermes profile gets its own model. Scout on DeepSeek. Analyst on Sonnet. Briefer on budget model. Coder on Opus. per-cron-job: pin a specific model to any cron job. morning brief on Haiku. deep research on Sonnet. monitoring on DeepSeek Flash. each job uses only the model it needs. per-session: /model deepseek/deepseek-v4-flash hot-swap mid-conversation. no restart needed. FALLBACK CHAINS: if your primary model is unavailable, Hermes automatically switches to the next provider. rate limit or server error = next model in the chain. no failed runs. no manual intervention. set in Desktop app, Dashboard, or config.yaml: fallback_providers: - openrouter - nous - codex PROVIDER PATHS: OPENROUTER: 300+ models under one API key. pay per token. most flexible. NOUS PORTAL: 300+ models + Tool Gateway (web search, image gen, TTS, browser). one OAuth. one subscription. 10% off token-billed providers. CHATGPT SUB: GPT-5.5 + Grok via OAuth. included tokens with $20 subscription. OLLAMA: free. local. private. zero API cost. your hardware only. mix providers across profiles and tasks. Scout on OpenRouter. Analyst on Nous Portal. Coder on ChatGPT sub. Monitor on Ollama. THE RULE: premium for work that needs deep reasoning. mid-range for daily driver tasks. budget for volume and background work. free for monitoring and routine jobs. pricing changes fast. check openrouter ai for current rates before committing. Which is your favourite model and for what task? full 15 levels breakdown in the article 👇

YanXbt

17,138 Aufrufe • vor 2 Monaten

I work on Gemini CLI as my day job and I've shipped features, shaped the roadmap, and helped organize many things behind the scenes for the last eight months or so. I hope I've made a difference for so many of our users. I love my day job, but even after hours I can't help but tinker. I bet you folks noticed? So here is (yet another!) experimental project. This isn't a replacement for any real products, but I had fun building it to see what I can do my my own. So why did I make Gemma CLI? Because I couldn't help but think... 1. What if the CLI was built around local models from the ground up? 2. What if I could think about how to correct for Gemma tool calling quirks? 3. What if I added a token-per-second counter, or showed reasoning events for those times the model just seems stuck? etc... So here we are. This isn't even my first mini agent harness CLI project, I've built a few more, they are on my GitHub but abandoned. What's next for this? Not sure if I'll ever open source this or even keep it around honestly... but feel free to give my experiment a try with the amazing Gemma models. Very lightly tested, so use at your own risk :-D

I work on Gemini CLI as my day job and I've shipped features, shaped the roadmap, and helped organize many things behind the scenes for the last eight months or so. I hope I've made a difference for so many of our users. I love my day job, but even after hours I can't help but tinker. I bet you folks noticed? So here is (yet another!) experimental project. This isn't a replacement for any real products, but I had fun building it to see what I can do my my own. So why did I make Gemma CLI? Because I couldn't help but think... 1. What if the CLI was built around local models from the ground up? 2. What if I could think about how to correct for Gemma tool calling quirks? 3. What if I added a token-per-second counter, or showed reasoning events for those times the model just seems stuck? etc... So here we are. This isn't even my first mini agent harness CLI project, I've built a few more, they are on my GitHub but abandoned. What's next for this? Not sure if I'll ever open source this or even keep it around honestly... but feel free to give my experiment a try with the amazing Gemma models. Very lightly tested, so use at your own risk :-D

Dmitry Lyalin

17,197 Aufrufe • vor 3 Monaten

New course to bring you up to state-of-the-art at using AI to help you code: Build Apps with Windsurf's AI Coding Agents, built in partnership with WIndsurf (Codeium) and taught by Anshul Ramachandran! AI-assisted IDEs (Integrated Development Environments) make developers’ workflows faster, more efficient, and much more fun. Agentic tools like Windsurf are more than just code autocomplete—they are collaborative coding agents that help you break down complex applications, iterate efficiently, and generate code that spans multiple files. Although a lot of coding assistants share the same underlying large language models for planning and reasoning, a major point of distinction is how they handle tools, keep track of context, and stay aligned with your intent as a developer. For instance, if you make modifications to a class definition in your code and make the same modifications to other classes in the same directory, you might tell the AI agent "Do the same thing in similar places in this directory." Here, tracking your intent means understanding that “the same thing" refers to that recent edit you just made, which must be followed by appropriate search and tool-calling to implement the changes. In this course, you'll learn the inner workings of coding agents, their strengths and limitations, and how to use Windsurf to quickly build several applications. In detail, you'll: - Build a mental model of how agents work by combining human-action tracking, tool integration, and context awareness to carry out an agentic coding workflow. - Learn the challenges of code search and discovery and how a multi-step retrieval approach helps coding agents address them. - Use Windsurf to analyze and understand a large, old codebase and update it to the latest versions of the frameworks and packages it uses. - Build a Wikipedia data analysis app that retrieves, parses, and analyzes word frequencies. - Enhance the performance of your Wikipedia analysis app by adding caching, and through this, also learn how to course-correct when the AI agent produces unexpected results. - Learn tips and tricks such as keyboard shortcuts, autocomplete, and @ mentions to quickly call on agentic capabilities. - Use image/multimodal capabilities of the AI agent to increase your development velocity; you'll see an example of uploading a mockup with sketched-out UI features, and ask the agent to use that to build new functionality to an app. By the end of this course, you’ll understand agentic coding in-depth and know how to use it to make your development process much faster, more efficient, and enjoyable. Please sign up here!

New course to bring you up to state-of-the-art at using AI to help you code: Build Apps with Windsurf's AI Coding Agents, built in partnership with WIndsurf (Codeium) and taught by Anshul Ramachandran! AI-assisted IDEs (Integrated Development Environments) make developers’ workflows faster, more efficient, and much more fun. Agentic tools like Windsurf are more than just code autocomplete—they are collaborative coding agents that help you break down complex applications, iterate efficiently, and generate code that spans multiple files. Although a lot of coding assistants share the same underlying large language models for planning and reasoning, a major point of distinction is how they handle tools, keep track of context, and stay aligned with your intent as a developer. For instance, if you make modifications to a class definition in your code and make the same modifications to other classes in the same directory, you might tell the AI agent "Do the same thing in similar places in this directory." Here, tracking your intent means understanding that “the same thing" refers to that recent edit you just made, which must be followed by appropriate search and tool-calling to implement the changes. In this course, you'll learn the inner workings of coding agents, their strengths and limitations, and how to use Windsurf to quickly build several applications. In detail, you'll: - Build a mental model of how agents work by combining human-action tracking, tool integration, and context awareness to carry out an agentic coding workflow. - Learn the challenges of code search and discovery and how a multi-step retrieval approach helps coding agents address them. - Use Windsurf to analyze and understand a large, old codebase and update it to the latest versions of the frameworks and packages it uses. - Build a Wikipedia data analysis app that retrieves, parses, and analyzes word frequencies. - Enhance the performance of your Wikipedia analysis app by adding caching, and through this, also learn how to course-correct when the AI agent produces unexpected results. - Learn tips and tricks such as keyboard shortcuts, autocomplete, and @ mentions to quickly call on agentic capabilities. - Use image/multimodal capabilities of the AI agent to increase your development velocity; you'll see an example of uploading a mockup with sketched-out UI features, and ask the agent to use that to build new functionality to an app. By the end of this course, you’ll understand agentic coding in-depth and know how to use it to make your development process much faster, more efficient, and enjoyable. Please sign up here!

Andrew Ng

139,978 Aufrufe • vor 1 Jahr

a team of researchers just proved you don't need a bigger model, you need a smarter plan researchers from Tsinghua and South China University of Technology built a framework called Atomic Task Graph. it turned 7B-8B open-source models into GPT-4 competitors on complex agent benchmarks, beating it on two out of three. no fine-tuning. no extra training. zero parameter updates. current AI agents plan in a straight line. step 1, step 2, step 3. when step 4 fails, the whole chain breaks. and the longer the chain gets, the more the model hallucinates because it's reasoning over a ballooning text history. here's how it works. 1. instead of a linear chain, ATG breaks any complex task into a directed graph where subtask inputs and outputs are explicitly mapped 2. it recursively decomposes each subtask until every node is one atomic tool call 3. independent branches run in parallel instead of waiting in line 4. before anything executes, a lightweight "thought experiment" simulates the plan internally to catch bad dependencies and missing steps early 5. when something breaks at runtime, ATG traces the failure to the exact subgraph that caused it and repairs only that piece. validated work stays frozen. the old way meant a failure at step 5 forced a full replan from scratch. hallucinated actions piled up the longer the task ran. ReAct hit a 43% hallucination rate on household tasks. ATG on an 8B Llama model scored 63.65 on ALFWorld. GPT-4 with ReAct scored 41.24 on the same benchmark. hallucinated actions dropped to 12%. those numbers happened because someone stopped throwing compute at the problem and started thinking about how work gets organized. that's the part that gets me. the industry is spending billions on scale. this team spent time on architecture. and the architecture won.

a team of researchers just proved you don't need a bigger model, you need a smarter plan researchers from Tsinghua and South China University of Technology built a framework called Atomic Task Graph. it turned 7B-8B open-source models into GPT-4 competitors on complex agent benchmarks, beating it on two out of three. no fine-tuning. no extra training. zero parameter updates. current AI agents plan in a straight line. step 1, step 2, step 3. when step 4 fails, the whole chain breaks. and the longer the chain gets, the more the model hallucinates because it's reasoning over a ballooning text history. here's how it works. 1. instead of a linear chain, ATG breaks any complex task into a directed graph where subtask inputs and outputs are explicitly mapped 2. it recursively decomposes each subtask until every node is one atomic tool call 3. independent branches run in parallel instead of waiting in line 4. before anything executes, a lightweight "thought experiment" simulates the plan internally to catch bad dependencies and missing steps early 5. when something breaks at runtime, ATG traces the failure to the exact subgraph that caused it and repairs only that piece. validated work stays frozen. the old way meant a failure at step 5 forced a full replan from scratch. hallucinated actions piled up the longer the task ran. ReAct hit a 43% hallucination rate on household tasks. ATG on an 8B Llama model scored 63.65 on ALFWorld. GPT-4 with ReAct scored 41.24 on the same benchmark. hallucinated actions dropped to 12%. those numbers happened because someone stopped throwing compute at the problem and started thinking about how work gets organized. that's the part that gets me. the industry is spending billions on scale. this team spent time on architecture. and the architecture won.

Alex Veremeyenko

173,351 Aufrufe • vor 1 Monat

Everyone is sleeping on Meta's SAM 3 release. But it's actually a big deal. Here's why: Companies spend millions paying humans to label images and videos frame by frame. A single autonomous driving dataset? Months of work, hundreds of annotators, millions in cost. Without labeled data, you can't train custom models. Without custom models, you're stuck with generic solutions. This is why most companies never move past pilots. SAM 3 breaks this cycle. First let's look at the evolution: SAM 1 segmented objects when you clicked on them. Revolutionary, but one object at a time. SAM 2 added video tracking with memory. Game-changing, but you still manually prompted every object. SAM 3 changes everything with text prompts. Type "yellow school bus" and it finds ALL of them in your image or video. Not just one. Every instance across thousands of frames. Now here's where people get confused: "Can't I just use GPT-5 or Gemini for this?" No, and here's why that's a terrible approach. Large multimodal LLMs are great for reasoning, but they're slow and expensive for production visual tasks. You're paying API costs per image, waiting seconds for responses, getting inconsistent results. SAM 3 runs in 30 milliseconds on a single GPU for 100+ objects. That's 100x faster, and you own the infrastructure. More importantly, SAM 3 gives you precise pixel-level masks, not descriptions. Try asking an LLM to segment every defective part on a manufacturing line in real-time. It won't work. SAM 3 does this effortlessly. The real breakthrough is their data engine. Meta built an AI-human hybrid system that's 5x faster for complex annotations. They trained SAM 3 on 4 million unique visual concepts - 50x more than existing benchmarks like LVIS. SAM 3 is trained on 4 million unique visual concepts, it handles everything: - Text-based concept search - Interactive refinement with clicks - Video tracking across frames - Zero-shot detection of new concepts The model is open source. Weights, code, and benchmarks are on GitHub. If you're building computer vision applications, this is the foundation model to evaluate. The annotation time savings alone will pay for integration costs within weeks. Find the relevant links in the next tweet!

Everyone is sleeping on Meta's SAM 3 release. But it's actually a big deal. Here's why: Companies spend millions paying humans to label images and videos frame by frame. A single autonomous driving dataset? Months of work, hundreds of annotators, millions in cost. Without labeled data, you can't train custom models. Without custom models, you're stuck with generic solutions. This is why most companies never move past pilots. SAM 3 breaks this cycle. First let's look at the evolution: SAM 1 segmented objects when you clicked on them. Revolutionary, but one object at a time. SAM 2 added video tracking with memory. Game-changing, but you still manually prompted every object. SAM 3 changes everything with text prompts. Type "yellow school bus" and it finds ALL of them in your image or video. Not just one. Every instance across thousands of frames. Now here's where people get confused: "Can't I just use GPT-5 or Gemini for this?" No, and here's why that's a terrible approach. Large multimodal LLMs are great for reasoning, but they're slow and expensive for production visual tasks. You're paying API costs per image, waiting seconds for responses, getting inconsistent results. SAM 3 runs in 30 milliseconds on a single GPU for 100+ objects. That's 100x faster, and you own the infrastructure. More importantly, SAM 3 gives you precise pixel-level masks, not descriptions. Try asking an LLM to segment every defective part on a manufacturing line in real-time. It won't work. SAM 3 does this effortlessly. The real breakthrough is their data engine. Meta built an AI-human hybrid system that's 5x faster for complex annotations. They trained SAM 3 on 4 million unique visual concepts - 50x more than existing benchmarks like LVIS. SAM 3 is trained on 4 million unique visual concepts, it handles everything: - Text-based concept search - Interactive refinement with clicks - Video tracking across frames - Zero-shot detection of new concepts The model is open source. Weights, code, and benchmarks are on GitHub. If you're building computer vision applications, this is the foundation model to evaluate. The annotation time savings alone will pay for integration costs within weeks. Find the relevant links in the next tweet!

Akshay 🚀

46,435 Aufrufe • vor 9 Monaten

Jensen Huang just BROKE the most important rule in the industry. And it explains why Nvidia controls 95% of the AI chip market. Last night at CES, he unveiled Vera Rubin - the new AI supercomputer that's shipping right now. Full production started weeks ago. But here's the part that made every semiconductor engineer in the room go crazy: Reuben GPU is 5x faster than Blackwell. But only has 1.6x the transistors. That should be physically impossible. Moore's Law says you get maybe 25% more performance per transistor generation. Jensen just delivered 300%. How? He BROKE the most sacred rule in chip design. The rule every company follows: "Never redesign more than 1-2 chips per generation." Nvidia redesigned all six chips simultaneously. Vera CPU. Reuben GPU. Connect X9 networking. Bluefield 4 DPU. MVLink switches. Spectrum X Ethernet. Every. Single. Component. From scratch. He calls it "extreme co-design." The industry calls it insane. One rack now moves 240 terabytes per second. That's TWICE the entire global internet bandwidth. In a single rack. And it runs on 45°C water - no chillers needed. Which saves 6% of global data center power. But the real story isn't the hardware... It's what they're doing with it. Nvidia just open-sourced Alpha Mayo. The world's first reasoning autonomous vehicle AI. Mercedes-Benz CLA launches with it in Q1. Europe Q2. Asia by year-end. Not a concept car. Not a limited release. Full production vehicles. And the AI will even explain its reasoning out loud. "I'm slowing down because the truck ahead is braking and there's a cyclist merging." It thinks. Then tells you what it's thinking. Then executes. Jensen drove it through San Francisco for an hour yesterday. No hands. No interventions. Through heavy Sunday traffic. The whole thing is open source now. Every line of training code. Every data source. The entire stack. But why would Nvidia give this away? Because they learned something from the last year: Open models activated the entire world. DeepSeek R1 proved open source can hit the frontier. Downloads exploded. Every country, every startup, every researcher can now build AI. And they all need Nvidia hardware to train it. That's the strategy. Give away the recipes. Sell the kitchen. The partnerships tell you where this is going: Siemens is integrating Nvidia into every industrial design tool. Cadence and Synopsys are rebuilding chip design around Nvidia. Palantir, ServiceNow, Snowflake - their entire platforms now run on Nvidia's agentic AI stack. This isn't just selling chips anymore. Nvidia is rebuilding the entire computing stack. From design to manufacturing to deployment. Every layer of the trillion-dollar AI infrastructure buildout runs through them. And now they're 18 months ahead of everyone else. Again. The competition is still trying to match Blackwell. Nvidia's already shipping the thing that makes Blackwell look slow. What do you think - is anyone catching them? The only company capable of this might be Google.

Jensen Huang just BROKE the most important rule in the industry. And it explains why Nvidia controls 95% of the AI chip market. Last night at CES, he unveiled Vera Rubin - the new AI supercomputer that's shipping right now. Full production started weeks ago. But here's the part that made every semiconductor engineer in the room go crazy: Reuben GPU is 5x faster than Blackwell. But only has 1.6x the transistors. That should be physically impossible. Moore's Law says you get maybe 25% more performance per transistor generation. Jensen just delivered 300%. How? He BROKE the most sacred rule in chip design. The rule every company follows: "Never redesign more than 1-2 chips per generation." Nvidia redesigned all six chips simultaneously. Vera CPU. Reuben GPU. Connect X9 networking. Bluefield 4 DPU. MVLink switches. Spectrum X Ethernet. Every. Single. Component. From scratch. He calls it "extreme co-design." The industry calls it insane. One rack now moves 240 terabytes per second. That's TWICE the entire global internet bandwidth. In a single rack. And it runs on 45°C water - no chillers needed. Which saves 6% of global data center power. But the real story isn't the hardware... It's what they're doing with it. Nvidia just open-sourced Alpha Mayo. The world's first reasoning autonomous vehicle AI. Mercedes-Benz CLA launches with it in Q1. Europe Q2. Asia by year-end. Not a concept car. Not a limited release. Full production vehicles. And the AI will even explain its reasoning out loud. "I'm slowing down because the truck ahead is braking and there's a cyclist merging." It thinks. Then tells you what it's thinking. Then executes. Jensen drove it through San Francisco for an hour yesterday. No hands. No interventions. Through heavy Sunday traffic. The whole thing is open source now. Every line of training code. Every data source. The entire stack. But why would Nvidia give this away? Because they learned something from the last year: Open models activated the entire world. DeepSeek R1 proved open source can hit the frontier. Downloads exploded. Every country, every startup, every researcher can now build AI. And they all need Nvidia hardware to train it. That's the strategy. Give away the recipes. Sell the kitchen. The partnerships tell you where this is going: Siemens is integrating Nvidia into every industrial design tool. Cadence and Synopsys are rebuilding chip design around Nvidia. Palantir, ServiceNow, Snowflake - their entire platforms now run on Nvidia's agentic AI stack. This isn't just selling chips anymore. Nvidia is rebuilding the entire computing stack. From design to manufacturing to deployment. Every layer of the trillion-dollar AI infrastructure buildout runs through them. And now they're 18 months ahead of everyone else. Again. The competition is still trying to match Blackwell. Nvidia's already shipping the thing that makes Blackwell look slow. What do you think - is anyone catching them? The only company capable of this might be Google.

Ricardo

580,166 Aufrufe • vor 7 Monaten