# Nightly Research Summary
Updated: 2026-09-05 07:17:16 AEST
Scope: bounded arXiv scan across cs.AI, cs.LG, cs.CR, and cs.RO.
## Latest Findings
- 2609.04199v1 - Compile by Training: Turning Natural-Language Specifications into Local Neural Functions (2026-09-03, cs.CL, cs.AI, cs.LG) - http://arxiv.org/abs/2609.04199v1 - Implication: high-impact autonomy gating.
- 2609.04198v1 - Clean Engineering, Unstable Measurement: A Preregistered Reliability Failure of Black-Box LLM Observers on Shared Endpoints (2026-09-03, cs.AI, cs.LG) - http://arxiv.org/abs/2609.04198v1 - Implication: watchlist only.
- 2609.04197v1 - ESPO: Error-Structured Prompt Optimization via Diagnose, Diversify, and Stabilize (2026-09-03, cs.CL, cs.AI) - http://arxiv.org/abs/2609.04197v1 - Implication: memory and retrieval governance.
- 2609.04194v1 - Legibility is Not Interpretability: Comparing Judged and Actual Importance in Chain-Of-Thought Reasoning (2026-09-03, cs.CL, cs.LG) - http://arxiv.org/abs/2609.04194v1 - Implication: watchlist only.
- 2609.04193v1 - GIFT: Guided Intermediate Feature Training via Action-Oriented Structural Supervision for Robotic Manipulation (2026-09-03, cs.RO) - http://arxiv.org/abs/2609.04193v1 - Implication: high-impact autonomy gating.
- 2609.04190v1 - One Editor, Many Edits: A Unified Training-Free Framework for Diverse Video Editing (2026-09-03, cs.CV, cs.AI) - http://arxiv.org/abs/2609.04190v1 - Implication: memory and retrieval governance.
- 2609.04189v1 - Robust PAC Learning of Concurrent Stochastic Games (2026-09-03, cs.LG, cs.GT, cs.LO, cs.MA) - http://arxiv.org/abs/2609.04189v1 - Implication: memory and retrieval governance.
- 2609.04183v1 - Seeing Before Synthesizing: VLM-Guided Transition Event Discovery for Weakly-Supervised Dense Video Captioning (2026-09-03, cs.CV, cs.AI) - http://arxiv.org/abs/2609.04183v1 - Implication: memory and retrieval governance.
Public mirror generated from the latest Hermes nightly arXiv research output. No private paths or local-only artifact locations are published.