Improving Large Vision and Language Models by Learning from a Panel of Peers

• URL: http://arxiv.org/abs/2509.01610v1
• Date: Mon, 01 Sep 2025 16:43:48 GMT
• Title: Improving Large Vision and Language Models by Learning from a Panel of Peers
• Authors: Jefferson Hernandez, Jing Shi, Simon Jenni, Vicente Ordonez, Kushal Kafle,
• Abstract summary: We propose a novel Panel-of-Peers learning framework inspired by collaborative learning among humans.<n>By simulating a peer review system, our models generate, assess, and refine outputs in response to a curated set of prompts.<n>Our experiments show significant improvement across multiple benchmarks, demonstrating the potential of peer evaluations as a scalable alternative to self-supervised alignment.
• Score: 27.83658413272528
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Traditional alignment methods for Large Vision and Language Models (LVLMs) primarily rely on human-curated preference data. Human-generated preference data is costly; machine-generated preference data is limited in quality; and self-supervised preference data often introduces hallucinations. To overcome these limitations, we propose a novel Panel-of-Peers learning framework inspired by collaborative learning among humans. This approach leverages a panel of LVLMs, each evaluating and learning from their collective outputs through an iterative self-improvement process. By simulating a peer review system, our models generate, assess, and refine outputs in response to a curated set of prompts, mimicking a classroom learning environment. We demonstrate that this methodology enhances model performance without requiring extensive human-labeled datasets. Our experiments show significant improvement across multiple benchmarks, demonstrating the potential of peer evaluations as a scalable alternative to self-supervised alignment. Notably, we show that Panel-of-Peers increases the average score on fifteen benchmarks from 48% to 57%

Related papers

• Measuring Teaching with LLMs [4.061135251278187]
  This paper uses custom Large Language Models built on sentence-level embeddings.<n>We show that these specialized models can achieve human-level and even super-human performance with expert human ratings above 0.65.<n>We also find that aggregate model scores align with teacher value-added measures, indicating they are capturing features relevant to student learning.
  arXiv  Detail & Related papers  (2025-10-27T03:42:04Z)
• No Labels Needed: Zero-Shot Image Classification with Collaborative Self-Learning [0.0]
  Vision-language models (VLMs) and transfer learning with pre-trained visual models appear as promising techniques to deal with this problem.<n>This paper proposes a novel zero-shot image classification framework that combines a VLM and a pre-trained visual model within a self-learning cycle.
  arXiv  Detail & Related papers  (2025-09-23T12:54:52Z)
• Improving Preference Extraction In LLMs By Identifying Latent Knowledge Through Classifying Probes [20.20764453136706]
  Large Language Models (LLMs) are often used as automated judges to evaluate text.<n>We propose using linear classifying probes, trained by leveraging differences between contrasting pairs of prompts, to access latent knowledge and extract more accurate preferences.
  arXiv  Detail & Related papers  (2025-03-22T12:35:25Z)
• VLFeedback: A Large-Scale AI Feedback Dataset for Large Vision-Language Models Alignment [55.7956150385255]
  We investigate the efficacy of AI feedback to scale supervision for aligning vision-language models. We introduce VLFeedback, the first large-scale vision-language feedback dataset. We train Silkie, an LVLM fine-tuned via direct preference optimization on VLFeedback.
  arXiv  Detail & Related papers  (2024-10-12T07:56:47Z)
• Self-Taught Evaluators [77.92610887220594]
  We present an approach that aims to im-proves without human annotations, using synthetic training data only. Our Self-Taught Evaluator can improve a strong LLM from 75.4 to 88.3 on RewardBench.
  arXiv  Detail & Related papers  (2024-08-05T17:57:02Z)
• Unpacking DPO and PPO: Disentangling Best Practices for Learning from Preference Feedback [110.16220825629749]
  Learning from preference feedback has emerged as an essential step for improving the generation quality and performance of modern language models. In this work, we identify four core aspects of preference-based learning: preference data, learning algorithm, reward model, and policy training prompts. Our findings indicate that all aspects are important for performance, with better preference data leading to the largest improvements.
  arXiv  Detail & Related papers  (2024-06-13T16:17:21Z)
• Silkie: Preference Distillation for Large Visual Language Models [56.10697821410489]
  This paper explores preference distillation for large vision language models (LVLMs) We first build a vision-language feedback dataset utilizing AI annotation. We adopt GPT-4V to assess the generated outputs regarding helpfulness, visual faithfulness, and ethical considerations. The resulting model Silkie, achieves 6.9% and 9.5% relative improvement on the MME benchmark regarding the perception and cognition capabilities.
  arXiv  Detail & Related papers  (2023-12-17T09:44:27Z)
• ULMA: Unified Language Model Alignment with Human Demonstration and Point-wise Preference [16.73260713938154]
  A typical alignment procedure consists of supervised fine-tuning and preference learning. We introduce Point-wise Direct Preference Optimization, a novel preference learning method designed to harness point-wise feedback effectively. Our work also uncovers a novel connection between supervised fine-tuning and point-wise preference learning, culminating in Unified Language Model Alignment.
  arXiv  Detail & Related papers  (2023-12-05T07:52:12Z)
• Improving Generalization of Alignment with Human Preferences through Group Invariant Learning [56.19242260613749]
  Reinforcement Learning from Human Feedback (RLHF) enables the generation of responses more aligned with human preferences. Previous work shows that Reinforcement Learning (RL) often exploits shortcuts to attain high rewards and overlooks challenging samples. We propose a novel approach that can learn a consistent policy via RL across various data groups or domains.
  arXiv  Detail & Related papers  (2023-10-18T13:54:15Z)
• Denoising and Selecting Pseudo-Heatmaps for Semi-Supervised Human Pose Estimation [38.97427474379367]
  We introduce a denoising scheme to generate reliable pseudo-heatmaps as targets for learning from unlabeled data. We select the learning targets from these pseudo-heatmaps guided by the estimated cross-student uncertainty. Our results show that our model outperforms previous state-of-the-art semi-supervised pose estimators.
  arXiv  Detail & Related papers  (2023-09-29T19:17:30Z)
• ASPEST: Bridging the Gap Between Active Learning and Selective Prediction [56.001808843574395]
  Selective prediction aims to learn a reliable model that abstains from making predictions when uncertain. Active learning aims to lower the overall labeling effort, and hence human dependence, by querying the most informative examples. In this work, we introduce a new learning paradigm, active selective prediction, which aims to query more informative samples from the shifted target domain.
  arXiv  Detail & Related papers  (2023-04-07T23:51:07Z)
• ILLUME: Rationalizing Vision-Language Models through Human Interactions [18.701950647429]
  We propose a tuning paradigm based on human interactions with machine-generated data. Our ILLUME executes the following loop: Given an image-question-answer prompt, the VLM samples multiple candidate rationales, and a human critic provides feedback via preference selection. This loop increases the training data and gradually carves out the VLM's rationalization capabilities that are aligned with human intent.
  arXiv  Detail & Related papers  (2022-08-17T11:41:43Z)

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.