Explore tools, prompts, skills, news and resources related to strong 第18页.
Prompt works related to strong 第18页
Latest articles, prompts, use cases and editorial content about strong 第18页
Simon Willison transcribed the compute section of Anthropic's IPO roadshow video, in which co-founder Tom Brown says Anthropic now trains and serves Claude across all four major chip families: AWS Trainium, Google T...
Researchers identify a failure mode called CLADE, where retrieval-augmented language models answer from passages that mention only a broader category rather than the specific answer required. On biomedical questions from...
Researchers analyzed 1,260,967 blind votes cast on Chatbot Arena between May 2023 and August 2024, restricting the analysis to pairs of models whose versions were fixed. For the same unchanged pair, the share of votes sa...
This arXiv paper analyzes why higher-order discretization does not generally improve the convergence order of stochastic diffusion samplers, establishing matching upper and worst-case lower bounds. It introduces forward-...
The paper proposes Falsification-Gated Self-Evolution (FGSE), a verifier-grounded framework for deciding when lessons derived from an agent episode should become persistent memory. FGSE represents each candidate lesson a...
Google DeepMind researchers used agentic AI and a framework called AlphaProof Code to formally verify key cryptographic guarantees of the Signal protocol in a four-week proof-of-concept. The work provides further indepen...
Anthropic claims its in-house model, Mythos, ported Claude to AMD GPUs in three days with almost no human intervention. The company says Claude now runs on four chip families, aiming to cut costs and reduce vendor depend...
The paper identifies memory-induced degradation in LLM agents, where retrieved lessons cause failures on tasks the agent could otherwise solve. It introduces WikiTrace, an evidence-linked memory system that stores each l...
Researchers propose ReFrain (Regression-Free Abstention), which combines a linear probe on first-token logits, threshold calibration on held-out image clusters, and a deterministic veto that replaces rejected outputs wit...
LM-SimBench is a unified framework, profiling dataset, and benchmark for evaluating LLM training simulators. It decomposes simulation into functional stages from model compilation to schedule-level aggregation and connec...
A review of 54 recent evaluation studies of ambient AI scribes in healthcare identifies ten evaluation dimensions and 40 distinct metrics, operationalized in 96 different ways. Thirty-five of those operationalizations, 3...
A ChinaTalk investigation reports that US data brokers including Mercor, Surge AI, and AfterQuery are selling roughly $500 million in AI training data to Chinese AI labs. The data reportedly includes reinforcement-learni...
The paper proposes Verifiable Agentic Traffic Class (VATC), which combines a reserved DSCP declaration for agent traffic, a router-side classifier that demotes flows contradicting the declaration, and a rate-limited prio...
An arXiv study evaluates 250 paired GSM8K derivations on Qwen3-8B and LLaMA-3.1-8B-Instruct, comparing clustered attention-head masking with budget-matched diffuse i.i.d. dropout. At onset B=0.10, Qwen3-8B accuracy fell...
This arXiv paper proposes an edge orchestration framework for deploying vision transformers across connected and automated vehicles. It jointly selects model partition points, per-layer quantization bitwidths, and execut...
This paper introduces AL-LBLP, an augmented Lagrangian method for solving Lifted Bellman Linear Programming in offline reinforcement learning. Instead of fixed hinge-penalty coefficients, it learns multipliers via dual a...
Researchers propose BARS, a sequential sampling rule for majority identification under an unknown distribution with a fixed budget and target error rate. BARS certifies the top-two gap and returns either an early stop, a...
Security firm Claroc found more than 35,000 Grok-built apps accessible to anyone without a login, including patient data and confidential SpaceXSI documents. The exposure stems from the "Anyone with the link" a...