arXiv AI

V-JEPA-LM: Video-Language Models That Read Dynamics Beyond Static, Text-Aligned Representations

The paper introduces V-JEPA-LM, a video‑language model that leverages the dynamic representations learned by V‑JEPA video encoders. By applying counterfactual supervision—pixel transformations that create answer‑balanced video variants—the model learns to read motion dynamics such as speed, direction, and order, rather than merely retrieving static attributes. Trained on 4.53 M public samples with 32 visual tokens per frame, V‑JEPA‑LM outperforms existing open‑source video LLMs across eleven benchmarks, with a 4 B model achieving a 72.1 average score.

arXiv Computer Vision
Oct 23

What Do Verifiable Rewards Teach Video-Language Models About Time? A Controlled Multi-Model Study

The paper investigates how reinforcement learning from verifiable rewards (RLVR) affects video‑language models’ understanding of time. Four open models were fine‑tuned under different data recipes—verified synthetic questions, unverified self‑supervised tasks, a 1:1 mix, and a verified+real mix—and evaluated on in‑domain and out‑of‑domain real‑video benchmarks with frames ordered, shuffled, or absent. Results show that verified training boosts in‑domain accuracy, largely through non‑visual gains, but can severely harm out‑of‑domain performance unless real verified data is added; moreover, no recipe induces genuine temporal‑order grounding, indicating that verifiable rewards improve benchmark scores without teaching temporal understanding.

By Avyay Sadhu, Patrick Cooper
arXiv Computer Vision
Oct 5

Behavior Pack Optimization for Video MLLM Post-Training

The paper introduces Behavior Pack Optimization (BPO), a post‑training method for video multimodal large language models that replaces single‑response rewards with a set of outputs across counterfactual views. BPO enforces stability when interventions are irrelevant, sensitivity when key evidence is removed, and abstention when no evidence remains, using an anchor‑relative advantage to keep the objective stable with small pack sizes. Experiments on datasets such as TempCompass, MVBench, and NExT‑QA show that BPO improves macro accuracy and abstention metrics for models like Qwen2.5‑VL‑7B‑Instruct, with gains that transfer to other benchmarks and models.

By Zhaolu Kang, Shiyu Liu, Tailong Luo, Wei Zhang, Yingjie He, Lei Wei, Guansu Wang, Liang He, Siheng Wang, Guangyuan Dong, Jiaqi Su, Shuang Chen, Haoyu Ji, Qishi Zhan, Kaiyue Zhou
arXiv Machine Learning
Jun 2

Perception First: A Frontier Native-Video Model with Self-Consistency for Implicit Video Question Answering

arXiv:2606. 01485v1 Announce Type: cross Abstract: We describe our submission to the VRR Challenge @ CVPR 2026, built on the \emph{ImplicitQA} / \emph{VRR-QA} benchmark~\cite{implicitqa}: multiple-choice video question answering in which answers are deliberately \emph{not} observable in any single frame and must be inferred from spatial layout, motion, depth, viewpoint, causality, and social context across discontinuous frames of creative video.

By Ali Alavi
arXiv Computer Vision
Sep 10

MotionBlind: Probing the Illusion of Motion Understanding in Video-LLMs

arXiv:2609.09528v1 Announce Type: new Abstract: Video large language models (Video-LLMs) are increasingly used as the perceptual front end of world models, a role that assumes they can read motion: h...

By Dhairya Bhatia, Bishoy Galoaa, Oliver Fritsche, Shahid Kamal, Muhammad Obaidullah Abdul Salam, Umer Saleem, Om Rastogi, Frania Felix Chettiar, Nesli Erdogmus, Sarah Ostadabbas
arXiv Computer Vision
Sep 18

SCOUT: Sim-to-Real Text-Based Person Retrieval by Embedding-Space Prediction over Frozen Video Features

SCOUT is a frozen‑encoder approach for sim‑to‑real text‑based person retrieval that predicts cross‑modal embeddings instead of fine‑tuning cross‑encoders. It uses a trainable predictor to map patch tokens from a frozen video encoder (V‑JEPA) into the embedding space of a frozen text encoder (EmbeddingGemma), guided by a bidirectional InfoNCE objective. The method achieves state‑of‑the‑art results on the AI City Challenge 2026 Track 4, with a full retrieve‑fuse‑rerank pipeline reaching 84.25 mAP@10 and a single frozen model alone scoring 60.63, while training costs are modest (≈95 GPU‑hours).

By Abdarahmane Traor\'e, Andy Couturier, \'Eric Hervet
arXiv Computer Vision
Oct 23

Ask About the Right Moment: Training-Free Object Identity Reasoning for Video-Language Models

The paper introduces a training‑free method called “Track, Then Bind” for video‑language models to reason about object identity over time. It separates semantic identity binding from temporal identity maintenance by using a promptable video tracker to follow candidate objects and a frozen vision‑language model to bind an object to a track based on visible evidence. The approach achieves state‑of‑the‑art performance on several benchmarks, outperforming fine‑tuned trajectory models and other training‑free agents on tasks such as VET‑Bench, Perception Test cup games, and TOC‑Bench.

By Di Yang, Xiaochang Li, Muyang Liu, Yufei Wang, Yanhai Xiong
arXiv Computation and Language
Nov 4

GAVAGAI: Referential Alignment for Infant-Scale Vision-Language Models via Balanced Optimal Transport

The paper introduces GAVAGAI, a method that adds a referential alignment variable to vision‑language models by solving a small assignment problem via entropic optimal transport. By allowing words to be assigned to a special NOTHING column and enforcing a capacity constraint on regions, the approach improves word‑to‑picture accuracy and localization on the SAYCam infant video corpus. Experiments show that the alignment loss boosts performance over a standard next‑token captioning baseline, especially under high non‑reference rates.

By Rajdeep Pal, Jotiraditya Banerjee, Saptarshi Pani
arXiv AI
Oct 9

Mid-Training Language Models on Raw Video

The paper investigates using raw, unlabeled video as a mid‑training phase for a pretrained language model. By encoding video frames into visual tokens and training the model to predict the next token, the authors mid‑train Qwen3‑1.7B on YT‑Temporal‑1B clips. The resulting model outperforms a non‑mid‑trained counterpart by 2.9 points on video benchmarks and 5.1 points on image benchmarks, while maintaining comparable text performance.

By Jaedong Hwang, Xiaoqian Shen, Ernie Chang, Changsheng Zhao, Chong Zhou, Saksham Suri, Qi Qian, Zechun Liu, Lemeng Wu, Qinsi Wang, Raghuraman Krishnamoorthi, Wei Wen
arXiv Computation and Language
Sep 25

Free the Language Model From the Vision Encoder: Semantic Serialization as a Perception Interface for Small Language Models

The paper introduces a perception interface that separates vision from language in vision‑language models. A frozen perception stack detects objects, a deterministic semantic serializer converts the perceived state into text, and a standard text‑only large language model (LLM) answers questions. Experiments on a campus‑robot benchmark show that this serialized interface outperforms a zero‑shot VLM of the same language‑model size, especially as the language model shrinks, and that the advantage persists under paraphrase and different supervision regimes.

By Cong Xu, Ravi Sankar