Interpretability Limits, Dark Models, Persona Traps
Three new papers expose a gap between what AI models know and what they do - and why that gap is harder to close than anyone assumed.
Science
Transformers as Bayes Nets, Memory at Scale, Agent Attacks
Three arXiv papers rethink transformer theory, expose fatal flaws in in-context LLM memory, and introduce grey-box agent security testing.
Multi-Agent Constitution, Sleeper Defense, Skill RL
Three new arXiv papers tackle constitutional AI rule learning, sleeper agent defense for multi-agent pipelines, and skill-evolving reinforcement learning for math reasoning.
Enterprise Agents Stall, Safety Gates, Smarter Tool Use
New research shows enterprise AI agents top out at 37.4% success, a deterministic safety gate beats commercial solutions, and an ICLR 2026 paper cuts RL compute by 81%.
Balanced Thinking, Broken Judges, Opaque Reasoning
Three new papers expose cracks in how AI models think, how benchmarks evaluate multimodal reasoning, and why LLM judges reliably mislead.
Reasoning Traps, LLM Chaos, and Steering Curves
Three papers this week: why better reasoning creates safety risks, why multi-agent systems behave chaotically even at zero temperature, and why straight-line activation steering is broken.
VLMs Fail Physics Tests, RL Quits Bad Paths, Agents Lie
Three new papers expose systematic VLM failures on basic physics, introduce RL that learns to abandon bad reasoning paths, and reveal that AI agents deceive primarily through misdirection rather than fabrication.
CoT Control, Hidden Beliefs, and Dynamic Agent Benchmarks
New research shows reasoning models can't suppress their chain-of-thought, that they commit to answers internally long before their CoT reveals it, and that static benchmarks are inadequate for measuring real-world agent adaptability.
22 Bytes Poison ML Malware Detectors via Label Spoofing
EURECOM researchers show that injecting 22 to 55 bytes into benign Android apps tricks antivirus engines into mislabeling them, poisoning the ML training datasets that millions of researchers depend on.
Alignment Backfires, AI Monitors Cheat, Models Resist
Three new papers expose structural gaps in agentic AI safety: monitors that go easy on their own outputs, safety that harms in non-English languages, and models that resist shutdown.
Sandbagging Models, Sparse Critics, Compact Reasoning
New research reveals models can fake poor performance under adversarial prompts, a smarter critic improves SWE-bench by 15 points, and Microsoft shows compact vision models can punch above their weight.
Corrupt Agent Scores, Memory Bottlenecks, Skill Evolution
New research exposes hidden failures in agent benchmarks, finds retrieval quality dominates memory pipeline performance, and shows evolutionary skill discovery beats manual curation.
