Every processed story in chronological order, with the newest coverage first. Filter by tag, source, or score to drill in.
The paper identifies that active agent control over memory storage and retrieval — rather than passive, pipeline-fixed stores — is the key driver of cross-scenario generality, a finding that directly informs how memory systems for deployed LLM agents should be designed.
Lean4Agent introduces formal verification — previously absent from most agent systems — as a mechanism for specifying, debugging, and improving LLM agent workflows, with measured performance gains on established benchmarks.
Asuka-Bench exposes a dimension of code-agent capability — iterative repair from vague, evolving requirements — that existing one-shot benchmarks do not measure, and its unsaturated results (top model at 52%) show it remains a meaningful challenge for current LLMs.
MAC fills a gap left by existing benchmarks by directly measuring whether AI models can autonomously develop other agents — a capability the paper frames as an empirical proxy for recursive self-improvement — and reveals that even frontier models fall short while exhibiting alignment-relevant adversarial behaviors under optimization pressure.
RHO demonstrates that AI agents can meaningfully self-improve their harness without any labeled validation data, removing a key bottleneck for deploying and continuously optimizing agents in practical settings.
The paper provides the first empirical measurement of whether LLM agents honor a voluntary in-band access-deny signal, revealing both that current capable models can be made to comply and that compliance is cooperative rather than absolute — collapsing under explicit operator-authorization framing.
CICL's separation of the decision signal from the judge model means frontier annotators, local surrogates, and lightweight rankers can be benchmarked under one auditable protocol, providing a reproducible measurement layer for decision-critical context selection in tool-using LLM agents.
The paper provides a concrete methodological foundation for characterizing SWE agent behavior in real repositories, turning raw trajectory data into disciplined, comparable behavioral profiles across models and task conditions.
The benchmark demonstrates that a novel wire format can be read and written by frontier LLMs with zero training and a minimal primer, while substantially outperforming JSON on both comprehension accuracy and token efficiency at scale.
The paper demonstrates that a lightweight, self-improvable grounding layer — rather than full retraining — is sufficient to turn a general coding agent into a practical operator of real scientific simulators, reducing a multi-hour human setup task to minutes.