AI Paper Atlas

Research map / Video / Video-language understanding

InfoTok: Adaptive Discrete Video Tokenizer via Information-Theoretic Compression

Haotian Ye, Qiyuan He, Jiaqi Han, Puheng Li, Jiaojiao Fan, Zekun Hao, Fitsum Reda, Yogesh Balaji, Huayu Chen, Sheng Liu, Angela Yao, James Zou and 3 more

ICLR 2026 · Video › Video-language understanding · approach: Long video understanding · 6 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use InfoTok instead of ElasticTok at BPP16 0.81tokenizer · TokenBench-256x256FVD: 141 -> 49Improved
use InfoTok instead of ElasticTok at BPP16 0.56tokenizer · TokenBench-256x256PSNR: 27.34 -> 29.27Improved
use InfoTok instead of ElasticTok at BPP16 0.56tokenizer · DAVIS-256x256FVD: 930 -> 540Improved
use InfoTok-Flex instead of InfoTok at BPP16 0.56tokenizer · DAVIS-256x256PSNR: 24.52 -> 24.84No clear effect
use ELBO-based router instead of exhaustive optimal search (ablation)InfoTok-Flex · TokenBench-256x256 BPP16 0.56PSNR: 29.39 -> 29.30No clear effect
use ELBO-based router instead of exhaustive optimal search (ablation)InfoTok-Flex · TokenBench-256x256 BPP16 0.31FVD: 165 -> 155No clear effect
use ELBO-based compressor instead of R2L masking (ablation)InfoTok · TokenBench BPP16 0.56PSNR: 27.43 -> 29.30Improved
use ELBO-based compressor instead of Jump masking (ablation)InfoTok · TokenBench BPP16 0.56FVD: 84 -> 71Improved
use ELBO-based compressor instead of Jump masking (ablation)InfoTok · TokenBench BPP16 0.56LPIPS: 0.173 -> 0.179Hurt
use ELBO adaptive mechanism instead of uniform on ViT backbone (ablation)ElasticTok ViT backbone · TokenBench BPP16 0.56FVD: 198 -> 114Improved
use InfoTok mechanism instead of ElasticTok mechanism on CosmosCosmos · inference latency per videolatency: 13.45 s -> 1.23 sImproved
use Cosmos + InfoTok at BPP16 0.56 instead of Cosmos at 1.00Cosmos Arch · TokenBench 360pFVD: 27 -> 56Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy