Every processed story in chronological order, with the newest coverage first. Filter by tag, source, or score to drill in.
RefGRPO demonstrates that agent self-assessment can be substantially improved without any external annotation or reward model, enabling agents to act as their own verifiers grounded in environment feedback.
HarnessX demonstrates that evolving the runtime scaffolding around a model — rather than scaling the model itself — can deliver substantial benchmark gains, offering a complementary path to agent improvement that does not require larger or more expensive models.
AgentSpec provides the first controlled compositional foundation for studying embodied LLM agents, revealing that scaffold interaction effects — not individual module quality — determine performance, which reframes how agent systems should be designed and compared.
Heterogeneous model pairs using tap recorded defects or requested changes in 69.8% of reviews versus 53.1% for homogeneous pairs, demonstrating that cross-vendor agent collaboration on a shared codebase produces broader code review coverage than single-vendor setups.
The paper establishes a fundamental, mathematically proven ceiling on multi-agent system performance that is determined by task structure — specifically C_min — meaning agent scaling and increased communication cannot overcome poorly structured tasks.
The benchmark reveals that dialogue capability is a distinct dimension of coding agent performance not captured by existing autonomous-system evaluations, exposing a gap between how agents are benchmarked and how they are actually used.
SecureClaw is the first architecture evaluated across AgentDojo, AgentLeak, and ASB in a common harness that closes both the plaintext-exposure and unauthorized-action boundaries simultaneously, rather than trading one surface for the other.
The framework replaces the standard text-concatenation bottleneck in multi-agent synthesis with direct KV cache consumption, cutting time-to-first-token by up to 11x while preserving or improving task accuracy across diverse benchmarks.
GTBP directly addresses the two key failure modes of existing context adaptation methods — inaccurate credit assignment and lack of convergence guarantees — in multi-LLM agentic pipelines, providing both theoretical stability proofs and empirical gains across three benchmarks.
DiffusionGemma's parallel token-generation architecture produces fluent but factually unreliable text, with error rates that grow as topics become more obscure — a concrete limitation that distinguishes it from its autoregressive counterpart for any fact-sensitive use case.