mlsanityData to 7 Oct 2026

Non-Linear Inference Time Intervention: Improving LLM Truthfulness

arXiv 2403.18680 ↗First posted 27 Mar 2024cs.CL cs.LGcited by 9

Not on the map. Why?

Abstract (arXiv)

In this work, we explore LLM's internal representation space to identify attention heads that contain the most truthful and accurate information. We further developed the Inference Time Intervention (ITI) framework, which lets bias LLM without the need for fine-tuning. The improvement manifests in introducing a non-linear multi-token probing and multi-token intervention: Non-Linear ITI (NL-ITI), which significantly enhances performance on evaluation benchmarks. NL-ITI is tested on diverse multiple-choice datasets, including TruthfulQA, on which we report over 16% relative MC1 (accuracy of model pointing to the correct answer) improvement with respect to the baseline ITI results. Moreover, we achieved a 10% relative improvement over the recently released Truth Forest (TrFf) method that also focused on ITI improvement.

Key facts. Cited by 9 papers in the map (data through 7 Oct 2026). Builds on 23 of its 26 references that are in the map. First posted on arXiv 27 Mar 2024.

Builds on

23 of 26 references.

References · 23

  1. 2022Training language models to follow instructions with human feedbackcited by 15,712
  2. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 13,958
  3. 2020Measuring Massive Multitask Language Understandingcited by 7,768
  4. 2018Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challengecited by 4,710
  5. 2019Parameter-Efficient Transfer Learning for NLPcited by 4,195
  6. 2021TruthfulQA: Measuring How Models Mimic Human Falsehoodscited by 2,919
  7. 2018Can a Suit of Armor Conduct Electricity? A New Dataset for Open Book Question Answeringcited by 2,114
  8. 2022Language Models (Mostly) Know What They Knowcited by 1,783
  9. 2023Inference-Time Intervention: Eliciting Truthful Answers from a Language Modelcited by 1,259
  10. 2023Representation Engineering: A Top-Down Approach to AI Transparencycited by 1,239
  11. 2023SelfCheckGPT: Zero-Resource Black-Box Hallucination Detection for Generative Large Language Modelscited by 787
  12. 2021BBQ: A Hand-Built Bias Benchmark for Question Answeringcited by 683
11 more
  1. 2023RLAIF vs. RLHF: Scaling Reinforcement Learning from Human Feedback with AI Feedbackcited by 615
  2. 2022ToxiGen: A Large-Scale Machine-Generated Dataset for Adversarial and Implicit Hate Speech Detectioncited by 419
  3. 2023HaluEval: A Large-Scale Hallucination Evaluation Benchmark for Large Language Modelscited by 329
  4. 2020Information-Theoretic Probing for Linguistic Structurecited by 204
  5. 2022On Second Thought, Let's Not Think Step by Step! Bias and Toxicity in Zero-Shot Reasoningcited by 191
  6. 2024Systematic Biases in LLM Simulations of Debatescited by 98
  7. 2023Do LLMs Possess a Personality? Making the MBTI Test an Amazing Evaluation for Large Language Modelscited by 54
  8. 2023Truth Forest: Toward Multi-Scale Truthfulness in Large Language Models through Intervention without Tuningcited by 45
  9. 2021A Non-Linear Structural Probecited by 31
  10. 2023Revisiting the Reliability of Psychological Scales on Large Language Modelscited by 16
  11. 2023Can We Use Probing to Better Understand Fine-tuning and Knowledge Distillation of the BERT NLU?cited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 5; 2026: 4.

9 papers.

  1. 2025The Scales of Justitia: A Comprehensive Survey on Safety Evaluation of LLMscited by 24
  2. 2025GrAInS: Gradient-based Attribution for Inference-Time Steering of LLMs and VLMscited by 6
  3. 2025LPASS: Linear Probes as Stepping Stones for vulnerability detection using compressed LLMscited by 4
  4. 2025Interpretable Risk Mitigation in LLM Agent Systemscited by 3
  5. 2025Distributional Surgery for Language Model Activationscited by 2
  6. 2026How LLMs Are Persuaded: A Few Attention Heads, Reroutedcited by 2
  7. 2026ORBIT: Training-Free Multi-Attribute Behavioral Steering via Orthogonal Subspace Rotationcited by 2
  8. 2026Multi-Adapter Representation Interventions via Energy Calibrationcited by 0
  9. 2026Adaptive Multi-Value Control in LLMs via Causal Activation Steeringcited by 0