Pith. sign in

REVIEW 2 cited by

AdaViP: Aligning Multi-modal LLMs via Adaptive Vision-enhanced Preference Optimization

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2504.15619 v1 pith:YI267IIE submitted 2025-04-22 cs.CV

AdaViP: Aligning Multi-modal LLMs via Adaptive Vision-enhanced Preference Optimization

classification cs.CV
keywords preferenceoptimizationvisualadaptivepreferencesadavipaligningalignment
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Preference alignment through Direct Preference Optimization (DPO) has demonstrated significant effectiveness in aligning multimodal large language models (MLLMs) with human preferences. However, existing methods focus primarily on language preferences while neglecting the critical visual context. In this paper, we propose an Adaptive Vision-enhanced Preference optimization (AdaViP) that addresses these limitations through two key innovations: (1) vision-based preference pair construction, which integrates multiple visual foundation models to strategically remove key visual elements from the image, enhancing MLLMs' sensitivity to visual details; and (2) adaptive preference optimization that dynamically balances vision- and language-based preferences for more accurate alignment. Extensive evaluations across different benchmarks demonstrate our effectiveness. Notably, our AdaViP-7B achieves 93.7% and 96.4% reductions in response-level and mentioned-level hallucination respectively on the Object HalBench, significantly outperforming current state-of-the-art methods.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Experience Augmented Policy Optimization for LLM Reasoning

    cs.LG 2026-06 conditional novelty 6.0

    EAPO injects a prior RL policy's token choices at critical decision points during rollout and beats standard RLVR baselines on math and science reasoning benchmarks.

  2. Experience Augmented Policy Optimization for LLM Reasoning

    cs.LG 2026-06 unverdicted novelty 5.0

    EAPO reuses prior RL policy experience adaptively at decision points in LLM rollouts with adapted importance sampling and reports gains over prior RLVR methods on math benchmarks.