Explicit Abstention Knobs for Predictable Reliability in Video Question Answering

Jorge Ortiz
12/31/2025
cs.AIcs.CV

No confident code implementation found for this paper.

Abstract

High-stakes deployment of vision-language models (VLMs) requires selective prediction, where systems abstain when uncertain rather than risk costly errors. We investigate whether confidence-based abstention provides reliable control over error rates in video question answering, and whether that control remains robust under distribution shift. Using NExT-QA and Gemini 2.0 Flash, we establish two findings. First, confidence thresholding provides mechanistic control in-distribution. Sweeping threshold epsilon produces smooth risk-coverage tradeoffs, reducing error rates f

View on arXivView PDF

Code Implementations

No confident code match yet

We couldn't find an author-owned or strongly-evidenced community implementation for this paper. Any repos shown below are weak matches — verify before relying on them.

No code implementations found yet.

Know of an implementation? Let us know in the comments below!

Similar Papers

Beyond Accuracy: On the Effects of Fine-tuning Towards Vision-Language Model's Prediction Rationality

82% match1 repo22024

Improving Selective Visual Question Answering by Learning from Your Peers

82% match1 repo262026

Prune-Then-Plan: Step-Level Calibration for Stable Frontier Exploration in Embodied Question Answering

82% match2025

Selective "Selective Prediction": Reducing Unnecessary Abstention in Vision-Language Reasoning

82% match1 repo32024

Learning Conformal Abstention Policies for Adaptive Risk Management in Large Language and Vision-Language Models

82% match1 repo62025

Improving Few-Shot Change Detection Visual Question Answering via Decision-Ambiguity-guided Reinforcement Fine-Tuning

81% match2025

Cite this paper

@article{ortiz2025explicit,
  title  = {Explicit Abstention Knobs for Predictable Reliability in Video Question Answering},
  author = {Jorge Ortiz},
  year   = {2025},
  eprint = {2601.00138},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2601.00138v2}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2601.00138.svg)](https://codeofpaper.com/paper/2601.00138)

Discussion