Inference-Time Intervention

Overview

Inference-time intervention studies how to steer a trained model while it is running, without necessarily changing its weights. Methods include activation and attention steering, decoding-time logit adjustments, retrieval-guided revision, context filtering, and reviewer agents that check tool calls or multi-step actions. They are applied to language and multimodal models, code generation, forecasting, theorem proving, and autonomous agents, with evaluations spanning safety, reasoning, factuality, and task success. A central challenge is improving the target behavior without harming useful outputs; intervention timing, persistence over long rollouts, robustness to new attacks, and reliable measurement remain open issues.

Latest papers

CardsList

37 papers4 papers in the last four weeks, up 33% on the four weeks before. 0.0% of all new papers.

  1. COTTI: Contrastive Training and Time-Aware Inference for Robust Industrial Recommendation under Promotional Drift

    Nov 17, 2026Maolei Huang, Jing Tian, Dexuan Duan +3Recommender SystemsLearning with Distribution Shift

  2. POEM: Output-Relative Block-Sparse Attention for Long-Context Prefilling

    Nov 10, 2026Haoyi Wu, Yunkai Dang, Wanyu Zhang +2LLM SafetyJailbreak Defense

  3. Whose Effect Is It? How Objectives Inflate Learned Subspace Interventions

    Nov 3, 2026Toygar Tanyel, Serkan AyvazMechanistic InterpretabilityExplainability Evaluation

  4. The First Few Tokens Matter: Understanding Chain-of-Thought Controllability

    Oct 29, 2026Hugh Van Deventer, Dmitrii Troitskii, Laura RuisCoT MonitoringInference-Time Intervention

  5. When Should We Intervene at Test Time?

    Oct 27, 2026Jiahao Zhang, Xinling Wen, Liangliang Li +1Inference-Time InterventionTest-Time Adaptation

  6. StepSafe: Learning Timely Intervention with Context-Aware Safety Reasoning

    Oct 14, 2026Siyi Gu, Kunhan Gao, Chao Peter Yang +3AI Agent SafetyLLM Agent Safety

  7. RolloutFaith: Auditing Persistent Internal Interventions in Visual World Model

    Sep 29, 2026Junchi Yao, Ziyi Wang, Youling Huang +1Inference-Time InterventionWorld Model Evaluation

  8. PASTABench: Proactive Assessment of Sequential Trajectories for Agent Safety

    Sep 23, 2026Jiapeng Sun, Yujin Zhou, Han Zhu +4AI Agent Safety BenchmarksInference-Time Intervention

  9. First Token Matters: Understanding Safety Collapse in Large Reasoning Models

    Sep 16, 2026Yizheng Yang, Haining Yu, Yuechen Wang +4LLM Safety AlignmentInference-Time Intervention

  10. Beyond Training: A Feasibility Taxonomy for Inference-Time AI Governance

    Sep 9, 2026Samar AnsariAI GovernanceInference-Time Intervention

  11. Mitigating LLM Over-Refusal via Dynamic Semantic Routing Calibration

    Sep 6, 2026Zixuan Wang, Bingjie Zhang, He Zhao +1Mechanistic InterpretabilityLLM Refusal Behavior

  12. Characterizing a Configuration Where Inference-Time PRM-Pruned Fragment Grafting Is Inert: Evidence from Three Reasoning LMs

    Sep 3, 2026Khawaja Murad ul Hassan, Mehran EbrahimiInference-Time InterventionLLM Reasoning

  13. OBJECTION! Lawyer Agents Mitigate Guilty Bias in Legal Judgment Prediction

    Sep 2, 2026Jaehoon Jeong, Jay-Yoon LeeLegal Judgment PredictionLegal Reasoning

  14. Low-Rank Dynamics-Effective Latent Carriers for Counterfactual Rollout in Learned World Models

    Aug 15, 2026Yang Liu, Yuming ChenWorld Model LearningLatent World Models

  15. Learned Interventions in Lean 4 grind

    Jul 25, 2026Evan Wang, Simon Chess, Sophie Szeto +1Automated Theorem ProvingLean Theorem Proving

  16. Uni-AdaVD: Universal Concept Erasure for Visual Generation via Orthogonal Value Decomposition

    Jul 16, 2026Qifan Zhou, Yuan Wang, Yanbin Hao +4Diffusion Model UnlearningInference-Time Intervention

  17. HyperSafe: Inference-Time Safety Recovery for Fine-Tuned Language Models

    Jul 13, 2026Aznaur Aliev, Carlos Hinojosa, Abdelrahman Eldesokey +3LLM Fine-TuningLLM Safety Alignment

  18. ScaleErasure: Inference-Time Minimal Intervention for Precise Concept Erasure in Next-Scale Autoregressive Image Generation

    Jun 28, 2026Cong Wang, Haiyu Wu, Zhiwei Jiang +4Image GenerationSafe T2I Generation

  19. DeSRPA: Decoupled Speech Role-Playing Agent via Inference-Time Intervention

    Jun 16, 2026Wenqiu Tang, Zhen Wan, Takahiro Komamizu +1Large Language Model-Based Role-Play SimulationSpoken Dialogue Systems

  20. Inference-Time Decision Calibration for Temporal Classification

    Jun 14, 2026Arthur Chagas, Arthur Buzelin, Yan Aquino +4Post-Hoc CalibrationTime Series Classification

  21. GITCO: Gated Inference-Time Context Optimization in TSFMs

    Jun 3, 2026Manya Pandey, Dhruv Kumar, Murari Mandal +1Zero-Shot Time Series ForecastingInference-Time Optimization

  22. Resonant Context Anchoring: Decoupling Attention Routing and Signal Gain at Inference Time

    Jun 1, 2026Mingkuan Zhao, Yide Gao, Wentao Hu +6Contextual FaithfulnessHallucination in Language Models

  23. Senses Wide Shut: A Representation-Action Gap in Omnimodal LLMs

    May 13, 2026Trung Nguyen Quang, Yiming Gao, Fanyi Pu +3Multimodal GroundingMultimodal Large Language Models

  24. SafeSteer: A Decoding-level Defense Mechanism for Multimodal Large Language Models

    May 12, 2026Xinyi Zeng, Xue Yang, Jingyuan Zhang +5Adversarial Attacks on VLMsMultimodal Robustness

  25. Defusing the Trigger: Plug-and-Play Defense for Backdoored LLMs via Tail-Risk Intrinsic Geometric Smoothing

    Apr 27, 2026Kaisheng Fan, Weizhe Zhang, Yishu Gao +2LLM Backdoor AttacksBackdoor Defense in LLMs