Pith. sign in

REVIEW 5 cited by

Reconstructive Visual Instruction Tuning

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

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 2410.09575 v2 pith:LC5PNUVP submitted 2024-10-12 cs.CV cs.AIcs.CLcs.LG

Haochen Wang , Anlin Zheng , Yucheng Zhao , Tiancai Wang , Zheng Ge , Xiangyu Zhang , Zhaoxiang Zhang This is my paper
classification cs.CVcs.AIcs.CLcs.LG
keywords visualrossimagesinputinstructionlmmsoutputssupervision
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

This paper introduces reconstructive visual instruction tuning (ROSS), a family of Large Multimodal Models (LMMs) that exploit vision-centric supervision signals. In contrast to conventional visual instruction tuning approaches that exclusively supervise text outputs, ROSS prompts LMMs to supervise visual outputs via reconstructing input images. By doing so, it capitalizes on the inherent richness and detail present within input images themselves, which are often lost in pure text supervision. However, producing meaningful feedback from natural images is challenging due to the heavy spatial redundancy of visual signals. To address this issue, ROSS employs a denoising objective to reconstruct latent representations of input images, avoiding directly regressing exact raw RGB values. This intrinsic activation design inherently encourages LMMs to maintain image detail, thereby enhancing their fine-grained comprehension capabilities and reducing hallucinations. Empirically, ROSS consistently brings significant improvements across different visual encoders and language models. In comparison with extrinsic assistance state-of-the-art alternatives that aggregate multiple visual experts, ROSS delivers competitive performance with a single SigLIP visual encoder, demonstrating the efficacy of our vision-centric supervision tailored for visual outputs.

Discussion (0). Sign in to comment.

Forward citations

Cited by 5 Pith papers

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

  1. Generation Enhances Understanding in Unified Multimodal Models via Multi-Representation Generation

    cs.CV 2026-01 conditional novelty 6.0 of 10

    Adding depth- and segmentation-generation objectives to UMM post-training improved spatial understanding and reduced hallucinations on Harmon and OpenUni while preserving generation quality.

  2. LaRe: Latent Refocusing for Multimodal Reasoning

    cs.CV 2025-11 reject novelty 6.0 of 10

    LaRe performs iterative visual refocusing in latent space and reports accuracy gains with fewer tokens, but its main experiments compare against baselines trained with less data.

  3. Reconstruction Alignment Improves Unified Multimodal Models

    cs.CV 2025-09 conditional novelty 6.0 of 10

    RECA, a self-supervised post-training objective that conditions unified multimodal models on their own visual understanding embeddings to reconstruct input images, improves text-to-image and editing benchmarks across ...

  4. SRUM: Fine-Grained Self-Rewarding for Unified Multimodal Models

    cs.CV 2025-10 conditional novelty 5.0 of 10

    A unified multimodal model can improve its own text-to-image generation by using its understanding module as a rewarder in a global-plus-local reward-weighted training loop.

  5. ReconVLA: Reconstructive Vision-Language-Action Model as Effective Robot Perceiver

    cs.RO 2025-08 unverdicted novelty 4.0 of 10

    Adding a reconstruction target that redraws the object region makes a vision-language-action model focus its attention on the right object and manipulate more precisely.

Pith tools