RefusalBench shows strict refusal rates fail to rank frontier LLMs correctly on biological safety, with provider effects and partial-compliance patterns that binary metrics miss.
The art of saying no: Contex- tual noncompliance in language models
8 Pith papers cite this work, alongside 2 external citations. Polarity classification is still indexing.
8
Pith papers citing it
2
external citations · Pith
abstract
Chat-based language models are designed to be helpful, yet they should not comply with every user request. While most existing work primarily focuses on refusal of "unsafe" queries, we posit that the scope of noncompliance should be broadened. We introduce a comprehensive taxonomy of contextual noncompliance describing when and how models should not comply with user requests. Our taxonomy spans a wide range of categories including incomplete, unsupported, indeterminate, and humanizing requests (in addition to unsafe requests). To test noncompliance capabilities of language models, we use this taxonomy to develop a new evaluation suite of 1000 noncompliance prompts. We find that most existing models show significantly high compliance rates in certain previously understudied categories with models like GPT-4 incorrectly complying with as many as 30% of requests. To address these gaps, we explore different training strategies using a synthetically-generated training set of requests and expected noncompliant responses. Our experiments demonstrate that while direct finetuning of instruction-tuned models can lead to both over-refusal and a decline in general capabilities, using parameter efficient methods like low rank adapters helps to strike a good balance between appropriate noncompliance and other capabilities.
representative citing papers
LLM responses to moral judgment queries reinforce implicit humanization, potentially exacerbating overreliance and misplaced trust.
Composable LoRA adapters can amplify or suppress OCEAN traits in LLMs, combine approximately additively, preserve moderate-scale capability, and move safety-relevant behaviours.
CPT is introduced as a pairwise reasoning-trace comparison stage that improves the reasoning-metacognition trade-off over standard SFT+RL pipelines across model scales.
Frontier LLMs exhibit premature closure by selecting answers at high rates on medical tasks where the correct choice was removed and on open-ended queries, with safety prompting reducing but not eliminating the behavior.
BAR trains independent domain experts via separate mid-training, SFT, and RL pipelines then composes them with a MoE router to match monolithic retraining performance at lower cost and without catastrophic forgetting.
Language models refuse 75.4% of requests to evade defeated rules and do so even after recognizing reasons that undermine the rule's legitimacy.
LLM safety evaluations are hindered by noise in dataset curation, automated red-teaming, response generation, and LLM-judge evaluation, making fair comparisons difficult and slowing progress.
citing papers explorer
-
RefusalBench: Why Refusal Rate Misranks Frontier LLMs on Biological Research Prompts
RefusalBench shows strict refusal rates fail to rank frontier LLMs correctly on biological safety, with provider effects and partial-compliance patterns that binary metrics miss.
-
Implicit Humanization in Everyday LLM Moral Judgments
LLM responses to moral judgment queries reinforce implicit humanization, potentially exacerbating overreliance and misplaced trust.
-
Persona Cartography: Charting Language Model Personality Traits in Weight Space
Composable LoRA adapters can amplify or suppress OCEAN traits in LLMs, combine approximately additively, preserve moderate-scale capability, and move safety-relevant behaviours.
-
Enhancing LLM Metacognition via Cognitive Pairwise Training
CPT is introduced as a pairwise reasoning-trace comparison stage that improves the reasoning-metacognition trade-off over standard SFT+RL pipelines across model scales.
-
Quantifying and Mitigating Premature Closure in Frontier LLMs
Frontier LLMs exhibit premature closure by selecting answers at high rates on medical tasks where the correct choice was removed and on open-ended queries, with safety prompting reducing but not eliminating the behavior.
-
Train Separately, Merge Together: Modular Post-Training with Mixture-of-Experts
BAR trains independent domain experts via separate mid-training, SFT, and RL pipelines then composes them with a MoE router to match monolithic retraining performance at lower cost and without catastrophic forgetting.
-
Blind Refusal: Language Models Refuse to Help Users Evade Unjust, Absurd, and Illegitimate Rules
Language models refuse 75.4% of requests to evade defeated rules and do so even after recognizing reasons that undermine the rule's legitimacy.
-
LLM-Safety Evaluations Lack Robustness
LLM safety evaluations are hindered by noise in dataset curation, automated red-teaming, response generation, and LLM-judge evaluation, making fair comparisons difficult and slowing progress.