arXiv:2611.03862v1 cs.LG

Dr. No: Feature Steering for Abstention in Vision-Language Models

Farid Hasanov, Riccardo Renzulli, Bahram Zonooz, Mykola Pechenizkiy, Decebal Constantin Mocanu

제출 2026-11-04

원문 초록

Large foundation models often answer even when the available visual or textual evidence is insufficient to support a reliable conclusion. This failure to abstain makes them hard to trust in high-stakes or information-seeking settings, where a confident but unsupported answer can be worse than no answer at all. It also raises a natural question: can the internal mechanisms behind this behaviour be identified and controlled at inference time? We present Dr. No, a feature-steering approach that mines sparse autoencoder (SAE) features associated with abstention and commitment, then modulates a small set of them during the prefill pass, without retraining the base model. Dr. No is evaluated on LLaVA-1.6 Mistral 7B across multimodal abstention benchmarks, including UNK-VQA and MM-UPD, and additionally benchmarked on language reasoning tasks using Qwen3.5-2B on AbstentionBench subsets. Against multiple uncertainty-based abstention baselines, feature steering achieves a more favourable balance between correct answering and appropriate abstention than post-hoc rejection methods. The gains are strongest on UNK-VQA and on MM-UPD cases involving image--question or question--answer incompatibility, and transfer most clearly to mathematical reasoning subsets of AbstentionBench. However, performance is more limited on absent-answer detection and retrieval-heavy multi-hop QA, such as MuSiQue. These results suggest that sparse internal features provide lightweight control handles for shaping uncertainty-aware behaviour in foundation models, while also showing that abstention control is source-dependent rather than governed by a single universal signal.

제출 이력

확인된 버전만 표시합니다. 전체 제출 이력은 arXiv 원문에서 확인하세요.

  1. arXiv:2611.03862v1제출 시각 미확인

함께 살펴볼 논문

공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.

공통 태그로 연결된 공개 논문이 아직 없습니다.

토론

이 논문과 연결된 포럼 글이 아직 없습니다.