Sycophancy in Vision-Language Models: A Systematic Analysis and an Inference-Time Mitigation Framework

• URL: http://arxiv.org/abs/2408.11261v2
• Date: Mon, 23 Jun 2025 03:00:38 GMT
• Title: Sycophancy in Vision-Language Models: A Systematic Analysis and an Inference-Time Mitigation Framework
• Authors: Yunpu Zhao, Rui Zhang, Junbin Xiao, Changxin Ke, Ruibo Hou, Yifan Hao, Ling Li,
• Abstract summary: We analyze sycophancy across vision-language benchmarks and propose an inference-time mitigation framework.<n>Our framework effectively mitigates sycophancy across all evaluated models, while maintaining performance on neutral prompts.
• Score: 18.54098084470481
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large Vision-Language Models (LVLMs) have shown significant capability in vision-language understanding. However, one critical issue that persists in these models is sycophancy, where models are unduly influenced by leading or deceptive prompts, resulting in biased outputs and hallucinations. Despite the rapid development of LVLMs, evaluating and mitigating sycophancy remains largely under-explored. In this work, we fill this gap by systematically analyzing sycophancy across multiple vision-language benchmarks and propose an inference-time mitigation framework. We curate leading queries and quantify the susceptibility of state-of-the-art LVLMs to prompt-induced bias, revealing consistent performance degradation and instability across models and tasks. Our analysis further uncovers model-specific behavioral traits, such as sentiment sensitivity and prediction polarity shifts under sycophancy. To mitigate these issues, we propose a training-free, model-agnostic framework that operates entirely at inference time. Our approach first employs a query neutralizer, leveraging an language model to suppress implicit sycophantic bias in user queries. We then introduce a sycophancy-aware contrastive decoding mechanism that dynamically recalibrates token-level output distributions by contrasting responses to neutralized and leading queries. Finally, an adaptive logits refinement module further modifies the contrasted logits by integrating both a adaptive plausibility filter and query sentiment scaler, ensuring coherent and robust generation. Extensive experiments demonstrate that this framework effectively mitigates sycophancy across all evaluated models, while maintaining performance on neutral prompts. Our results suggest that sycophancy in LVLMs is a general and urgent challenge, and that inference-time strategies offer a promising path toward trustworthy multimodal reasoning.

Related papers

• On Evaluating Performance of LLM Inference Serving Systems [11.712948114304925]
  We identify recurring anti-patterns across three key dimensions: Baseline Fairness, Evaluation setup, and Metric Design.<n>These anti-patterns are uniquely problematic for Large Language Model (LLM) inference due to its dual-phase nature.<n>We provide a comprehensive checklist derived from our analysis, establishing a framework for recognizing and avoiding these anti-patterns.
  arXiv  Detail & Related papers  (2025-07-11T20:58:21Z)
• Seeing is Believing? Mitigating OCR Hallucinations in Multimodal Large Language Models [22.43132625619281]
  We propose KIE-HVQA, the first benchmark dedicated to evaluating OCR hallucination in degraded document understanding.<n>This dataset includes test samples spanning identity cards and invoices, with simulated real-world degradations for OCR reliability.<n>Experiments on Qwen2.5-VL demonstrate that our 7B- parameter model achieves a 22% absolute improvement in hallucination-free accuracy over GPT-4o.
  arXiv  Detail & Related papers  (2025-06-25T06:44:07Z)
• Flattery in Motion: Benchmarking and Analyzing Sycophancy in Video-LLMs [6.527988482383714]
  Video large language models (Video-LLMs) are increasingly integrated into real-world applications that demand grounded multimodal reasoning.<n>Sycophancy, the tendency of these models to align with user input even when it contradicts the visual evidence, undermines their trustworthiness in such contexts.<n>We propose VISE (Video-LLM Sycophancy Benchmarking and Evaluation), the first dedicated benchmark designed to evaluate sycophantic behavior in state-of-the-art Video-LLMs.
  arXiv  Detail & Related papers  (2025-06-08T15:00:21Z)
• Supervised Optimism Correction: Be Confident When LLMs Are Sure [91.7459076316849]
  We establish a novel theoretical connection between supervised fine-tuning and offline reinforcement learning.<n>We show that the widely used beam search method suffers from unacceptable over-optimism.<n>We propose Supervised Optimism Correction, which introduces a simple yet effective auxiliary loss for token-level $Q$-value estimations.
  arXiv  Detail & Related papers  (2025-04-10T07:50:03Z)
• Model Hemorrhage and the Robustness Limits of Large Language Models [119.46442117681147]
  Large language models (LLMs) demonstrate strong performance across natural language processing tasks, yet undergo significant performance degradation when modified for deployment.<n>We define this phenomenon as model hemorrhage - performance decline caused by parameter alterations and architectural changes.
  arXiv  Detail & Related papers  (2025-03-31T10:16:03Z)
• Attention Reallocation: Towards Zero-cost and Controllable Hallucination Mitigation of MLLMs [62.9348974370985]
  We propose attention reallocation (AttnReal) to mitigate hallucinations with nearly zero extra cost. Our approach is motivated by the key observations that, MLLM's unreasonable attention distribution causes features to be dominated by historical output tokens. Based on the observations, AttnReal recycles excessive attention from output tokens and reallocates it to visual tokens, which reduces MLLM's reliance on language priors.
  arXiv  Detail & Related papers  (2025-03-11T11:52:37Z)
• VisFactor: Benchmarking Fundamental Visual Cognition in Multimodal Large Language Models [62.667142971664575]
  We introduce VisFactor, a novel benchmark derived from the Factor-Referenced Cognitive Test (FRCT) VisFactor digitalizes vision-related FRCT subtests to systematically evaluate MLLMs across essential visual cognitive tasks. We present a comprehensive evaluation of state-of-the-art MLLMs, such as GPT-4o, Gemini-Pro, and Qwen-VL.
  arXiv  Detail & Related papers  (2025-02-23T04:21:32Z)
• Causality can systematically address the monsters under the bench(marks) [64.36592889550431]
  Benchmarks are plagued by various biases, artifacts, or leakage.<n>Models may behave unreliably due to poorly explored failure modes.<n> causality offers an ideal framework to systematically address these challenges.
  arXiv  Detail & Related papers  (2025-02-07T17:01:37Z)
• Mitigating Hallucination for Large Vision Language Model by Inter-Modality Correlation Calibration Decoding [66.06337890279839]
  Large vision-language models (LVLMs) have shown remarkable capabilities in visual-language understanding for downstream multi-modal tasks. LVLMs still suffer from generating hallucinations in complex generation tasks, leading to inconsistencies between visual inputs and generated content. We propose an Inter-Modality Correlation Decoding (IMCCD) method to mitigate hallucinations in LVLMs in a training-free manner.
  arXiv  Detail & Related papers  (2025-01-03T17:56:28Z)
• Towards More Robust Retrieval-Augmented Generation: Evaluating RAG Under Adversarial Poisoning Attacks [45.07581174558107]
  Retrieval-Augmented Generation (RAG) systems have emerged as a promising solution to mitigate hallucinations.<n>RAG systems are vulnerable to adversarial poisoning attacks, where malicious passages injected into the retrieval corpus can mislead models into producing factually incorrect outputs.<n>We present a rigorously controlled empirical study of how RAG systems behave under such attacks and how their robustness can be improved.
  arXiv  Detail & Related papers  (2024-12-21T17:31:52Z)
• Sycophancy in Large Language Models: Causes and Mitigations [0.0]
  Large language models (LLMs) have demonstrated remarkable capabilities across a wide range of natural language processing tasks. Their tendency to exhibit sycophantic behavior poses significant risks to their reliability and ethical deployment. This paper provides a technical survey of sycophancy in LLMs, analyzing its causes, impacts, and potential mitigation strategies.
  arXiv  Detail & Related papers  (2024-11-22T16:56:49Z)
• CATCH: Complementary Adaptive Token-level Contrastive Decoding to Mitigate Hallucinations in LVLMs [74.36850397755572]
  CATCH addresses issues related to visual defects that cause diminished fine-grained feature perception and cumulative hallucinations in open-ended scenarios. It is applicable to various visual question-answering tasks without requiring any specific data or prior knowledge, and generalizes robustly to new tasks without additional training.
  arXiv  Detail & Related papers  (2024-11-19T18:27:31Z)
• Mitigating Hallucinations of Large Language Models in Medical Information Extraction via Contrastive Decoding [92.32881381717594]
  We introduce ALternate Contrastive Decoding (ALCD) to solve hallucination issues in medical information extraction tasks. ALCD demonstrates significant improvements in resolving hallucination issues compared to conventional decoding methods.
  arXiv  Detail & Related papers  (2024-10-21T07:19:19Z)
• Have the VLMs Lost Confidence? A Study of Sycophancy in VLMs [44.56018149475948]
  sycophancy is a prevalent hallucination that poses significant challenges to visual language models (VLMs) We propose a synthetic dataset for training and employ methods based on prompts, supervised fine-tuning, and DPO to mitigate sycophancy. Our findings indicate that the ability to prevent sycophancy is predominantly observed in higher layers of the model.
  arXiv  Detail & Related papers  (2024-10-15T05:48:14Z)
• Mitigating Hallucinations in Large Vision-Language Models (LVLMs) via Language-Contrastive Decoding (LCD) [13.430637580980164]
  Large Vision-Language Models (LVLMs) are an extension of Large Language Models (LLMs) that facilitate processing both image and text inputs, expanding AI capabilities. Our study introduces a Language Contrastive Decoding (LCD) algorithm that adjusts LVLM outputs based on Large Language Models distribution confidence levels. Our method effectively improves LVLMs without needing complex post-processing or retraining, and is easily applicable to different models.
  arXiv  Detail & Related papers  (2024-08-06T08:10:34Z)
• Self-Introspective Decoding: Alleviating Hallucinations for Large Vision-Language Models [30.26685485474035]
  Large Vision-Language Models (LVLMs) have rapidly advanced in recent years. The prevalent issue known as the hallucination' problem has emerged as a significant bottleneck. We propose a simple yet effective method named Self-Introspective Decoding (SID)
  arXiv  Detail & Related papers  (2024-08-04T13:50:17Z)
• CODE: Contrasting Self-generated Description to Combat Hallucination in Large Multi-modal Models [51.70129969269271]
  We introduce a novel contrastive-based decoding method, COuntering DEscription Contrastive Decoding (CODE) Our method significantly reduces hallucinations and improves cross-modal consistency across various benchmarks and cutting-edge LMMs.
  arXiv  Detail & Related papers  (2024-06-04T03:04:21Z)
• VALOR-EVAL: Holistic Coverage and Faithfulness Evaluation of Large Vision-Language Models [57.43276586087863]
  Large Vision-Language Models (LVLMs) suffer from hallucination issues, wherein the models generate plausible-sounding but factually incorrect outputs. Existing benchmarks are often limited in scope, focusing mainly on object hallucinations. We introduce a multi-dimensional benchmark covering objects, attributes, and relations, with challenging images selected based on associative biases.
  arXiv  Detail & Related papers  (2024-04-22T04:49:22Z)
• Debiasing Multimodal Large Language Models [61.6896704217147]
  Large Vision-Language Models (LVLMs) have become indispensable tools in computer vision and natural language processing. Our investigation reveals a noteworthy bias in the generated content, where the output is primarily influenced by the underlying Large Language Models (LLMs) prior to the input image. To rectify these biases and redirect the model's focus toward vision information, we introduce two simple, training-free strategies.
  arXiv  Detail & Related papers  (2024-03-08T12:35:07Z)
• IBD: Alleviating Hallucinations in Large Vision-Language Models via Image-Biased Decoding [37.16880672402059]
  Over-reliance on linguistic priors has been identified as a key factor leading to hallucinations. We propose to alleviate this problem by introducing a novel image-biased decoding technique. Our method derives the next-token probability distribution by contrasting predictions from a conventional LVLM with those of an image-biased LVLM.
  arXiv  Detail & Related papers  (2024-02-28T16:57:22Z)
• Measuring and Improving Chain-of-Thought Reasoning in Vision-Language Models [61.28463542324576]
  Vision-language models (VLMs) have recently demonstrated strong efficacy as visual assistants that can generate human-like outputs. We evaluate existing state-of-the-art VLMs and find that even the best-performing model is unable to demonstrate strong visual reasoning capabilities and consistency. We propose a two-stage training framework aimed at improving both the reasoning performance and consistency of VLMs.
  arXiv  Detail & Related papers  (2023-09-08T17:49:44Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.