LLM Agents Can Easily Tamper With Their Own Traces. It centres on Claude Code, and also names Grok. Reported by arXiv. Bharat Hunt files it under AI Models and AI Regulation — the section covering a new or updated model, its capabilities, benchmarks or availability.
Written by Bharat Hunt from the headline and the coverage below. The original reporting is the source of truth.
Coding Agents for Generalized Task and Motion Planning Problems. It centres on Claude Code, and also names GPT and Benchmarks. Reported by arXiv. Bharat Hunt files it under AI Coding and AI Models — the section covering code generation, developer agents, IDEs and software engineering.
SkillGym: Internalizing Human Skills into LLMs for Real-World Problem Solving. It centres on Inference, and also names Fine-tuning and Claude Code. Reported by arXiv. Bharat Hunt files it under AI Models and AI Coding — the section covering a new or updated model, its capabilities, benchmarks or availability.
CliffCompaction: Cost-Efficient Compaction for Long-Horizon Coding Agents. It centres on GPT, and also names Claude Code. Reported by arXiv. Bharat Hunt files it under AI Coding and AI Models — the section covering code generation, developer agents, IDEs and software engineering.
GameLogicBench: Evaluating Coding Agents on Runtime Game Logic with Tick-Level State Assertions. It centres on Benchmarks, and also names Claude Code. Reported by arXiv. Bharat Hunt files it under AI Coding and AI Models — the section covering code generation, developer agents, IDEs and software engineering.