arXiv:2608.08446v2 Announce Type: replace Abstract: Personalized generation systems retrieve user history by request–memory relevance and inject it into…
MemWM: Memory-Augmented Text-Based World Model
arXiv:2608.07107v2 Announce Type: replace Abstract: World models are increasingly used to support planning in agents by predicting how environment states…
Towards Query-Agnostic RAG Evaluation via Query Coverage and Claim Verifiability
arXiv:2608.11238v2 Announce Type: replace Abstract: Retrieval-augmented generation improves the factuality of large language models by grounding responses…
Fragility of Value under Imperfect Alignment
arXiv:2607.28881v4 Announce Type: replace Abstract: As more responsibility is placed upon AI systems, it becomes increasingly important to guarantee that…
The Illusion of Equivalency: Statistical Characterization of Quantization Effects in LLMs
arXiv:2607.08734v2 Announce Type: replace Abstract: Post-Training Quantization has become widely used to compress large language models to make them…
Share the Judge, Learn the Deferral: Where Specialization Helps LLM Evaluation
arXiv:2607.27984v2 Announce Type: replace Abstract: Agentic systems generate outputs faster than human review. We contrast two LLM evaluator…
When Words Are Safe But Actions Kill: Probing Physical Jailbreak Beyond Textual Jailbreak in Hidden-State Risk Space
arXiv:2607.15218v2 Announce Type: replace Abstract: Large language models (LLMs) increasingly serve as high-level planners for embodied agents, where…
WorldLines: Benchmarking and Modeling Long-Horizon Stateful Embodied Agents
arXiv:2606.18847v2 Announce Type: replace Abstract: To assist humans over extended periods in real homes, embodied agents must remember user routines,…
AgRefactor: Self-Evolving Agentic Workflow for HLS Compatibility and Performance
arXiv:2606.30949v2 Announce Type: replace Abstract: High-Level Synthesis (HLS) provides a fast path from concepts to silicon, but converting real-world…
AI News Brief Hourly Summary 2026-08-25 03h : 13 posts
13 posts published in the last hour 00:32SafeSteer: Localized On-Policy Distillation for Efficient Safety Alignment 00:32Can LLMs Introspect? A Reality Check 00:32Self-Revising Discovery Systems for Science: A Categorical Framework for Agentic Artificial Intelligence 00:32Weak Critics Make Strong Learners: On-Policy Critique…
SafeSteer: Localized On-Policy Distillation for Efficient Safety Alignment
arXiv:2606.02530v2 Announce Type: replace Abstract: Aligning Large Language Models (LLMs) with human values often degrades their general capabilities,…
Can LLMs Introspect? A Reality Check
arXiv:2605.26242v2 Announce Type: replace Abstract: Can large language models detect and report their own internal states? A number of recent studies have…
Self-Revising Discovery Systems for Science: A Categorical Framework for Agentic Artificial Intelligence
arXiv:2606.01444v2 Announce Type: replace Abstract: Scientific discovery is not only answer generation but revision of the representational regime in…
Weak Critics Make Strong Learners: On-Policy Critique Distillation for Scalable Oversight
arXiv:2606.00424v2 Announce Type: replace Abstract: As large language models become stronger, weak supervisors may fail to provide reliable labels,…
Situational Awareness, star AI hedge fund that nearly imploded, now being probed by the SEC
The AI hedge fund went from “the talk of Wall Street” to “subject of federal subpoenas” faster than you can say “diversify your portfolio.”
GRASP: Gated Regression-Aware Skill Proposer for Self-Improving LLM Agents
arXiv:2605.29668v3 Announce Type: replace Abstract: LLM agents acting in structured environments fail in operational rather than conversational ways, and…
SEISMO: Explanation-Aware, Trajectory-Conditioned LLM Agents for Sample-Efficient Molecular Optimisation
arXiv:2602.00663v3 Announce Type: replace Abstract: Optimizing molecules to achieve desired properties is a central bottleneck across the chemical…
Two Heads are Better Than One: Test-time Scaling of Multi-agent Collaborative Reasoning
arXiv:2504.09772v3 Announce Type: replace Abstract: Test-Time Scaling has emerged as a powerful method to extend the reasoning capabilities of Large…