abliterationrefusal removalpublished Nov 16, 2026 · arXiv:2611.17846
Is Abliteration Getting Harder? Not Uniformly: Refusal Removal by Direction-Based Ablation and ARA in Heretic on the Qwen3.8, Gemma 4, Olmo 3.1 and Mistral Small 3.2 LLMs Versus Their Predecessors
factual reliabilityfalse-answer adoptionpublished Nov 16, 2026 · arXiv:2611.17763
Knowledge After Collapse: Context and Wording Shape Factual Accessibility in Language Models
medical imagingvision-language modelspublished Nov 11, 2026 · arXiv:2611.13006
Seeing Is Not Believing: Counterfactual Lesion Erasure Teaches a Small Medical Vision-Language Agent to Report Only What the Image Shows
language modelscalibrated decision probabilitiespublished Oct 31, 2026 · arXiv:2610.37732
ThinkJev: One Thought, Every Budget for Calibrated Typed Decisions
weakening-induced on-policy self-distillationlow-bit quantizationpublished Oct 31, 2026 · arXiv:2610.38681
Weaken to Strengthen: Weakening-Induced On-Policy Self-Distillation for Large Reasoning Models
LLM agentsharness environmentspublished Oct 15, 2026 · arXiv:2610.14118
ClawLoop: Self-Improving Agents via Harness-Native Training Loop