Pith. sign in

REVIEW 5 cited by

SlowFast-LLaVA-1.5: A Family of Token-Efficient Video Large Language Models for Long-Form Video Understanding

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 2503.18943 v2 pith:V7FUB7I2 submitted 2025-03-24 cs.CV

Mingze Xu , Mingfei Gao , Shiyu Li , Jiasen Lu , Zhe Gan , Zhengfeng Lai , Meng Cao , Kai Kang
show 2 more authors
Yinfei Yang Afshin Dehghan
This is my paper · ORCID
classification cs.CV
keywords videounderstandinglong-formmodelsresultssf-llava-1achievesfamily
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

We introduce SlowFast-LLaVA-1.5 (abbreviated as SF-LLaVA-1.5), a family of video large language models (LLMs) offering a token-efficient solution for long-form video understanding. We incorporate the two-stream SlowFast mechanism into a streamlined training pipeline, and perform joint video-image training on a carefully curated data mixture of only publicly available datasets. Our primary focus is on highly efficient model scales (1B and 3B), demonstrating that even relatively small Video LLMs can achieve state-of-the-art performance on video understanding, meeting the demand for mobile-friendly models. Experimental results demonstrate that SF-LLaVA-1.5 achieves superior performance on a wide range of video and image tasks, with robust results at all model sizes (ranging from 1B to 7B). Notably, SF-LLaVA-1.5 achieves state-of-the-art results in long-form video understanding (e.g., LongVideoBench and MLVU) and excels at small scales across various video benchmarks.

Discussion (0). Continue with ORCID 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. Towards Temporal Compositional Reasoning in Long-Form Sports Videos

    cs.CV 2026-04 unverdicted novelty 7.0 of 10

    SportsTime plus Chain-of-Time Reasoning (temporal-reward GRPO and anchor-observe-infer) modestly lifts open-ended sports VideoQA and step-wise temporal grounding over 4B–8B MLLM baselines.

  2. Stateful Token Reduction for Long-Video Hybrid VLMs

    cs.CV 2026-02 conditional novelty 6.0 of 10

    For hybrid Mamba–Transformer video models, keeping 25% of visual tokens with a query-based progressive schedule gives 3.8–4.2x prefilling speedups with near-baseline accuracy; the paper attributes this to stateful com...

  3. UniGen: Enhanced Training & Test-Time Strategies for Unified Multimodal Understanding and Generation

    cs.CV 2025-05 conditional novelty 6.0 of 10

    UniGen shows a 1.5B model trained on open data can beat larger systems on image understanding and generation once it verifies its own outputs with chain-of-thought and Best-of-N selection.

  4. FlexSelect: Flexible Token Selection for Efficient Long Video Understanding

    cs.CV 2025-06 conditional novelty 5.0 of 10

    FlexSelect selects a small fraction of query-relevant visual tokens using attention from an intermediate layer, improving long-video accuracy and inference speed across multiple VideoLLMs.

  5. Context-Aware Multi-Turn Visual-Textual Reasoning in LVLMs via Dynamic Memory and Adaptive Visual Guidance

    cs.CV 2025-09 reject novelty 3.0 of 10

    The proposed CAMVR framework is not supported by verifiable evidence, and the manuscript itself labels its experimental results as fabricated.

Pith tools