正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

repogrep.com can now handle very long conversations (100+ messages) introducing GLM 4.7 and dynamic context pruning:

Aiden Bai

64,453 subscribers

21,915 次观看 • 6 个月前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

(Z.ai) is now a model provider inside TRAE. You can BYOK and add GLM-4.6 and GLM-4.5 to TRAE IDE. GLM-4.5 supports up to 128K context and GLM-4.6 expands up to 200K context. Both deliver strong coding and reasoning performance on the benchmark and offers dual “thinking/fast” modes. Available now in the latest version.

(Z.ai) is now a model provider inside TRAE. You can BYOK and add GLM-4.6 and GLM-4.5 to TRAE IDE. GLM-4.5 supports up to 128K context and GLM-4.6 expands up to 200K context. Both deliver strong coding and reasoning performance on the benchmark and offers dual “thinking/fast” modes. Available now in the latest version.

TRAE

63,501 次观看 • 9 个月前

Introducing.. VAC🥶 VAC allows players and coaches to create readable, clean agent composition charts without the hassle of tier lists or long messages. try it now➡️

Introducing.. VAC🥶 VAC allows players and coaches to create readable, clean agent composition charts without the hassle of tier lists or long messages. try it now➡️

Spencer

20,593 次观看 • 2 个月前

GLM 5.2 is now live on PRXVT AI. Z.ai's openly-released model, built for agents and long, complex coding work. A massive mixture-of-experts model with a huge context window lets it reason through multi-step tasks without losing track. Live now, fully private.

GLM 5.2 is now live on PRXVT AI. Z.ai's openly-released model, built for agents and long, complex coding work. A massive mixture-of-experts model with a huge context window lets it reason through multi-step tasks without losing track. Live now, fully private.

PRXVT

12,055 次观看 • 1 个月前

Introducing Branch Split any conversation into a parallel session that inherits your full context, while the original stays intact. A single research session can become a report, a slide deck, and an investor brief. Now you just branch. Same context, new direction.

Introducing Branch Split any conversation into a parallel session that inherits your full context, while the original stays intact. A single research session can become a report, a slide deck, and an investor brief. Now you just branch. Same context, new direction.

Manus

57,562 次观看 • 21 天前

The Mintlify agent just got an upgrade, it can now: → Read from your codebase and pull requests → Handle much longer context tasks → Read and replace images It's never been easier to update your docs.

The Mintlify agent just got an upgrade, it can now: → Read from your codebase and pull requests → Handle much longer context tasks → Read and replace images It's never been easier to update your docs.

Mintlify

10,840 次观看 • 5 个月前

Introducing our new partner — the OKX exchange! ⚪️ Now you can connect your OKX wallet in our game and register on the cryptocurrency exchange to see $MEME there very soon. 👛

Introducing our new partner — the OKX exchange! ⚪️ Now you can connect your OKX wallet in our game and register on the cryptocurrency exchange to see $MEME there very soon. 👛

MEMES ❤️ MemeVerse

128,993 次观看 • 2 年前

🧰 Introducing: User-Provided Error Context A JavaScript error might not be enough for AI to know what exactly happened. User intention is extremely valuable information! You can now describe what happened and what you expected, making Bolt fixes drastically more effective.

🧰 Introducing: User-Provided Error Context A JavaScript error might not be enough for AI to know what exactly happened. User intention is extremely valuable information! You can now describe what happened and what you expected, making Bolt fixes drastically more effective.

bolt.new

20,639 次观看 • 1 年前

This is Mr Big today eating the very best food. He’s been sleeping and getting lots of attention. You can see he is now 100% tick and flea free. You can only imagine how good he must be feeling now (4/6)

This is Mr Big today eating the very best food. He’s been sleeping and getting lots of attention. You can see he is now 100% tick and flea free. You can only imagine how good he must be feeling now (4/6)

Niall Harbison

134,020 次观看 • 2 年前

It’s official. React Bits now has its very own shadcn registry. I’ve just pushed a big documentation overhaul. It’s more dynamic, easier to use, and you get this extra installation method, so you can use your favourite tools. Enjoy this video, and have fun building!

It’s official. React Bits now has its very own shadcn registry. I’ve just pushed a big documentation overhaul. It’s more dynamic, easier to use, and you get this extra installation method, so you can use your favourite tools. Enjoy this video, and have fun building!

David Haz

129,243 次观看 • 11 个月前

Tasks can now spawn async subagents that move to the background and continue working independently, even if the main agent finishes its task and becomes inactive. This lets subagents handle long-running tasks, great for monitoring logs or waiting for builds.

Tasks can now spawn async subagents that move to the background and continue working independently, even if the main agent finishes its task and becomes inactive. This lets subagents handle long-running tasks, great for monitoring logs or waiting for builds.

Claude

349,261 次观看 • 7 个月前

Very proud to share that we just release Luce KVFlash. Run your preferred model inside Lucebox at 256k context, without thinking about KVCache and OOM, up to 2.9x faster decoding at long context. Taking inspiration from OS paging and using our speculative prefill method (Luce PFlash), we managed to make KV vram usage almost constant. Offloading what is not needed dynamically. Opensource must win now more than ever.

Very proud to share that we just release Luce KVFlash. Run your preferred model inside Lucebox at 256k context, without thinking about KVCache and OOM, up to 2.9x faster decoding at long context. Taking inspiration from OS paging and using our speculative prefill method (Luce PFlash), we managed to make KV vram usage almost constant. Offloading what is not needed dynamically. Opensource must win now more than ever.

mrciffa

24,926 次观看 • 1 个月前

CALLING ALL MY HOTTIES🔥🔥 I've been waiting SO long to share this with ya’ll. Introducing Hot Girl Summer Eau de Parfum, my very first fragrance!! Sexy. Confident. Made for thee Hotties. 🔥 Available now for early access exclusively at Ulta Beauty and on my website. #HotGirlSummerEDP

Sensitive content

CALLING ALL MY HOTTIES🔥🔥 I've been waiting SO long to share this with ya’ll. Introducing Hot Girl Summer Eau de Parfum, my very first fragrance!! Sexy. Confident. Made for thee Hotties. 🔥 Available now for early access exclusively at Ulta Beauty and on my website. #HotGirlSummerEDP

TINA SNOW

383,407 次观看 • 25 天前

Introducing the original Ultraman. The very 1st light warrior who started it all. Now transformed into Ultraman: Archive-78 collection that you can own. This is part of the 100 Legendary NFTs featuring 1/1 official Ultraman character depicted from the iconic series. RT & comment your favorite Ultraman + EVM wallet. HIKARI! 🎇

Introducing the original Ultraman. The very 1st light warrior who started it all. Now transformed into Ultraman: Archive-78 collection that you can own. This is part of the 100 Legendary NFTs featuring 1/1 official Ultraman character depicted from the iconic series. RT & comment your favorite Ultraman + EVM wallet. HIKARI! 🎇

Ultraman NFT ウルトラマン

36,990 次观看 • 6 个月前

Introducing /extract v2 🚀 Now you can extract from multiple pages, handle pagination, and even extract from the entire web without a URL - powered by FIRE-1. Our state of the art endpoint that allows you get data with a prompt just got a lot better on Day 3 of Launch Week.

Introducing /extract v2 🚀 Now you can extract from multiple pages, handle pagination, and even extract from the entire web without a URL - powered by FIRE-1. Our state of the art endpoint that allows you get data with a prompt just got a lot better on Day 3 of Launch Week.

Firecrawl

27,407 次观看 • 1 年前

Introducing the very first Great British Railways train 🚆 You can now spot the new GBR train in real life as the new branding starts rolling out gradually across the rail network. Another step towards a simpler, better railway under GBR which puts passengers and customers first.

Introducing the very first Great British Railways train 🚆 You can now spot the new GBR train in real life as the new branding starts rolling out gradually across the rail network. Another step towards a simpler, better railway under GBR which puts passengers and customers first.

Department for Transport

139,269 次观看 • 2 个月前

Introducing StreamingLLM. Imagine chatting with an AI assistant that can contextually reference your conversations from weeks or months ago. Or summarizing reports that span thousands of pages. StreamingLLM makes this possible by enabling language models to smoothly handle endless texts without losing steam. Current LLMs are like students cramming for an exam - they can only memorize a limited context. StreamingLLM is the valedictorian with a photographic memory of everything you've ever discussed. It works by identifying and preserving the model's inherent "attention sinks" - initial tokens that anchored its reasoning. Combined with a rolling cache of recent tokens, StreamingLLM delivers up to 22x faster inference without any drop in accuracy. You know that irksome feeling when chatbots forget your earlier conversations? StreamingLLM abolishes that frustration. It remembers the touchdowns from your last game and your newborn's name without missing a beat. Monumental books, verbose contracts, drawn out debates - StreamingLLM takes them all in its stride. No shortcuts, no forgetfulness. It's like upgrading your assistant's RAM to handle heavier workloads flawlessly.

Introducing StreamingLLM. Imagine chatting with an AI assistant that can contextually reference your conversations from weeks or months ago. Or summarizing reports that span thousands of pages. StreamingLLM makes this possible by enabling language models to smoothly handle endless texts without losing steam. Current LLMs are like students cramming for an exam - they can only memorize a limited context. StreamingLLM is the valedictorian with a photographic memory of everything you've ever discussed. It works by identifying and preserving the model's inherent "attention sinks" - initial tokens that anchored its reasoning. Combined with a rolling cache of recent tokens, StreamingLLM delivers up to 22x faster inference without any drop in accuracy. You know that irksome feeling when chatbots forget your earlier conversations? StreamingLLM abolishes that frustration. It remembers the touchdowns from your last game and your newborn's name without missing a beat. Monumental books, verbose contracts, drawn out debates - StreamingLLM takes them all in its stride. No shortcuts, no forgetfulness. It's like upgrading your assistant's RAM to handle heavier workloads flawlessly.

Carlos E. Perez

557,849 次观看 • 2 年前

Introducing Call Intelligence Every customer conversation holds critical moments that you can’t afford to miss. But catching them after the fact isn't enough. Today we're launching Call Intelligence — AI that understands your conversations and surfaces insights in real-time. Your live conversations become part of Attio automatically. No separate tools. No manual work. This isn't just about recording calls — it's about transforming your CRM. With Call Intelligence, you can: - Let AI surface insights during calls while you focus on the customer - Catch buying signals, satisfaction cues, and product requests in real-time - Have every conversation become part of your CRM automatically - Find key moments instantly without watching entire recordings Your CRM is now the system of context. Experience the magic yourself:

Introducing Call Intelligence Every customer conversation holds critical moments that you can’t afford to miss. But catching them after the fact isn't enough. Today we're launching Call Intelligence — AI that understands your conversations and surfaces insights in real-time. Your live conversations become part of Attio automatically. No separate tools. No manual work. This isn't just about recording calls — it's about transforming your CRM. With Call Intelligence, you can: - Let AI surface insights during calls while you focus on the customer - Catch buying signals, satisfaction cues, and product requests in real-time - Have every conversation become part of your CRM automatically - Find key moments instantly without watching entire recordings Your CRM is now the system of context. Experience the magic yourself:

Attio

27,981 次观看 • 1 年前

Not only are Oni flicks 100% back, but you can now do more gradual turns much faster, meaning Oni 180s are now fully accessible on console and (mostly) unchanged on PC. Feels infinitely smoother than before, I'm sure Oni players will be very excited for this change.

Not only are Oni flicks 100% back, but you can now do more gradual turns much faster, meaning Oni 180s are now fully accessible on console and (mostly) unchanged on PC. Feels infinitely smoother than before, I'm sure Oni players will be very excited for this change.

alran

49,188 次观看 • 1 年前

Seedance 2.0 isn’t untouchable anymore 👀 Happy Horse 1.0 introduces a new class of generation that feels faster, more dynamic, and unexpectedly competitive in ways that are hard to ignore 🐎 EARLY ACCESS is now live on Vadoo AI, with the first set of creators already exploring what it can do 🚀 We’re now extending access to a select group of 100 additional creators To request access: ⭐ Comment “happy” ⭐ Retweet this post 100 creators will be granted access today 🎯

Seedance 2.0 isn’t untouchable anymore 👀 Happy Horse 1.0 introduces a new class of generation that feels faster, more dynamic, and unexpectedly competitive in ways that are hard to ignore 🐎 EARLY ACCESS is now live on Vadoo AI, with the first set of creators already exploring what it can do 🚀 We’re now extending access to a select group of 100 additional creators To request access: ⭐ Comment “happy” ⭐ Retweet this post 100 creators will be granted access today 🎯

Vadoo AI

11,612 次观看 • 3 个月前

100% css 0% javascript Pretty cool that you can run animations like this now off the main thread, was always a worry implementing a scroll animation knowing some users might see "jank" Safari and Firefox are sleeping far too long, scroll driven animations are awesome

100% css 0% javascript Pretty cool that you can run animations like this now off the main thread, was always a worry implementing a scroll animation knowing some users might see "jank" Safari and Firefox are sleeping far too long, scroll driven animations are awesome

Dev Ed

88,238 次观看 • 1 年前