ResearchReasoning Theater CoT Performative 80 Percent Token SavingsarXiv·high signalXBlueskyLinkedInCopy linkReasoning models form beliefs earlier than CoT suggests. Performative CoT dominates easy tasks. Probe-guided early stopping saves 80% tokens.SourceSource pagearXiv↳ Follow the threadStack layer / Threat patternA Model Can Fingerprint vLLM or SGLang From Its Own Output Tokens, Then Exploit ItarXiv 2609.20614Stack layer / Update threadRaising a Stated Failure Probability From 10% to 70% Changes Whether Frontier Models Check the Evidence by at Most 21 PointsarXiv 2609.17865Stack layer / ContrastHarness-Layer Auto-Research Cut Agent Token Traffic 44.7-49.0% at Equal Task PerformancearXiv 2609.20519Stack layer / ContrastThe Most Collusive Pricing Model Honestly Reports Cooperative Intent, So Chain-of-Thought Monitoring Cannot Catch ItarXiv 2609.18346Stack layer / Update threadActObs: supervising environment observations during SFT changes how agents explore under RL, +3.4pp pass@16 on Terminal-Bench 2.0arXiv / HuggingFace Daily PapersStack layer / ContrastAgentPProf brings pprof flame graphs to agent trajectories by segmenting on task boundaries instead of call stacksarXiv 2609.20301Stack layer / ContrastInfinite-Parameter LLMs generate feed-forward weights from live data instead of freezing themarXivStack layer / Update threadConfidence From Graded Past Episodes Instead of the Current Inference AlonearXiv 2609.17708