A new study uses five frontier LLMs from Anthropic and OpenAI as 'agentic curators' in a self-contained workspace to automate phenotype annotation. The agents achieved consistency within the range of human curators and substantially outperformed traditional NLP tools, addressing the scalability bottleneck in ontology curation.

This paper introduces Orthogonal Concept Erasure (OCE), which uses multiplicative parameter updates for precise concept removal while preserving generative capacity, supporting multi-concept erasure with high speed.

This paper empirically evaluates LLM-generated reviews for scientific papers, finding limited alignment with human reviews that varies significantly across prompts and models. It also shows that authors can game the system by iteratively revising papers based on LLM feedback, achieving statistically significant score increases for up to 35% of papers.

The Cognitive Categorical Transformer (CCT) is a 306M-parameter architecture that augments GPT-2 Small with cognitive and category-theoretic components, achieving 21.27 perplexity on WikiText-103, a 2.92 (12%) reduction over a fine-tuned baseline. Ablations attribute 84% of the improvement to GT-Full simplicial message passing. The study also identifies a structure/consistency distinction among categorical priors.

This paper proposes behavior-aware auxiliary corrections to stabilize off-policy temporal-difference learning. By replacing the auxiliary covariance matrix with the behavior Bellman matrix, the authors introduce BA-TDC and BA-TDRC algorithms. Theoretical analysis proves fixed-point preservation and almost-sure convergence. Experiments on standard benchmarks show that the behavior-aware replacement improves performance, but regularization is needed for robust results.

This paper proposes STHTD-MP, a behavior-induced Mirror-Prox temporal-difference method that replaces the covariance metric with the symmetric part of the behavior-policy Bellman matrix to improve off-policy prediction speed. Theoretical convergence analysis and numerical experiments on several benchmarks show improved performance over GTD2-MP.

Anthropic announced in its $65 billion Series H funding that its annualized run-rate revenue crossed $47 billion in early May 2026, up from $30 billion in April and $14 billion in February. The rapid growth has drawn comparisons to unprecedented organic revenue scaling, though some skeptics question the numbers. Anecdotal evidence of a client spending $500 million in a single month on Claude licenses adds context.

Together AI built the fastest speech-to-text stack on Artificial Analysis by treating ASR as a full-path systems problem, not just a GPU inference problem. This article details optimizations including TensorRT multi-profile encoders, conditional CUDA graphs, shared memory, evented I/O, and gc.freeze() to eliminate tail latency.

This article explores the practical application of reinforcement learning in post-training large language models, highlighting that the current bottleneck is infrastructure rather than algorithms. Modal shares its experience running RL post-training at scale and introduces its open-source library to help teams address key challenges like multi-node training, environment management, and GPU utilization.

This is the first post in the Profiling in PyTorch series, starting with a simple matrix multiplication and bias addition to teach readers how to use torch.profiler. It covers setting up the profiler, reading the profiler table and trace, understanding CPU/GPU activity gaps, and the impact of warmup and matrix size on performance regimes.

Anthropic released Claude Opus 4.8, described as a modest but tangible improvement over its predecessor. Key highlights include enhanced honesty (reduced unsupported claims, four times less likely to overlook code flaws), and new features like mid-conversation system messages. Pricing remains unchanged, but fast mode costs are significantly reduced.

Release of llm-anthropic 0.25.1 adds support for Claude Opus 4.8, fast mode option for eligible accounts, and changes default max_tokens to each model's maximum output.