正在加载视频...
视频加载失败
Introducing: A native HyperEVM explorer with analytics and dev tools for CoreWriter, precompiles, HIP-3, HIP-4, and more. HyperCore and the HyperEVM share the same state. This reads both sides of it. Real-Time: every page updates the moment a block lands. Home, blocks, transactions, CoreWriter, any address. You never need... show more
0 条评论
暂无评论
原始帖子的评论将显示在这里
![For the Sui Overflow 2026 hackathon, I built Talise. A gasless dollar account that sends money by name, settles in under a second, and keeps the amount private on-chain. This is what "money that moves like a message" actually looks like when you build it on the right rails. The core experience is simple by design. Sign in with Google/Apple, get a self-custodial wallet via zkLogin, claim your handle like sele@talise, and start sending. No seed phrase. No gas token. No 0x address to copy and verify. The complexity lives in the infrastructure. The user never sees it. [Kostas Kryptos] Sending is just a name. Type a handle, enter an amount, confirm. It arrives in under a second. Stablecoin transactions on Sui cost nothing, so what you type is exactly what arrives. No spread taken on the send. No fee waiting at the destination. Talise also has a built in privacy feature. The amount is hidden on-chain and the connection between sender and recipient is completely broken. Nobody watching the chain can see what moved or who it moved between. Real privacy, live on Sui mainnet today, not a line on a roadmap. Beyond the core send: claim links let you wrap dollars in a URL and drop it in any chat, no address needed to receive. Payment streaming lets you send a salary or a subscription by the second. Idle balances route automatically to on-chain lending so your dollars are never dead weight between payments. And cash-out to your local bank is live now in Nigeria, with more corridors coming. Every transaction is sponsored through a gas station so users never hold or spend a gas token. zkLogin, sponsored transactions, sub-second finality, and programmable transaction blocks are what made this possible. Sui is the only chain where this experience exists without compromise. Over 1,400 handles claimed on-chain. Every feature shipped and working in the live build. Testflight app: Talise. Money that moves like a message. Sui Community Official sarah 😾](https://image.24vids.com/tw-2068819361763074064/amplify_video_thumb/2068811614078259200/img/zq9Mn9_SdMGaJwAM.jpg)
![Self Attention by hand ✍️ ~ 9 steps walkthrough below Self-attention is what enables LLMs to understand context. How does it work? So I drew and calculated one entirely by hand. Goal: turn four 6D features into four 3D attention weighted features, filling in every cell yourself. = 1. Given = Four feature vectors, six dimensions each, one per position. = 2. Query, key, value = Let us multiply the features by WQ, WK and WV. Queries, keys and values all come out of the same four features, and that is what the word "self" is doing in self-attention. = 3. Prepare for MatMul = We copy the queries across the top and the transposed keys down the side. Lining the two up is half the work. = 4. MatMul = Let us multiply K transpose by Q. Every cell is the dot product of one key with one query, which we use as a matching score. That works because the dot product is the numerator of cosine similarity: it is how alike two vectors are, before anyone divides by their lengths. = 5. Scale = We divide by the square root of dk, the dimension of a key vector, here 3. Without it the scores grow with the dimension and a 64-wide head would swamp the softmax. To keep the page doable in pen, the drawing approximates dividing by root 3 with halving. = 6. e to the power = Let us raise e to the power of each score. This is the first half of softmax, and the drawing uses 3 in place of e, which is close enough to do in your head. = 7. Sum = We add up each column: 16, 6, 7 and 12. = 8. Normalize = Let us divide every cell by its column sum. That gives the attention weight matrix in yellow, and each of its four columns is now a probability distribution over the four positions. The decimals are nudged as they are rounded, so every column still sums to exactly 1. = 9. MatMul = We multiply the value vectors by those weights. Each output is a blend of all four values, mixed in the proportion the attention matrix just decided, and it goes to the position-wise feed forward network in the next layer: the FFN box at the bottom of the page. The outputs: Attention weights (A), by column = [.2, .6, 0, .2], [.2, .4, .2, .2], [.4, .2, 0, .4], [.1, .7, .1, .1] Attention weighted features (Z) = [8, 2, 6], [8, 4, 4], [16, 4, 2], [4, 2, 7] The takeaway: attention is a weighted average, and everything before step 9 exists to decide the weights. Compare every position with every other, turn the scores into one distribution per position, then blend. 💾 Save this post!](https://image.24vids.com/tw-2079933315990057242/ext_tw_video_thumb/2079933302048219136/pu/img/q3wr-ymUWG3Ai-ij.jpg)