AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

KATANA: Knowledge-Aligned Topology-Aware Neural Agents for RL-Driven Vision-Language Model Compression

Nafew Azim, Mir Ali, Fuad Rahman, Nabeel Mohammed

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Attention-based visual token pruning · 0 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 cubic sparsity schedule instead of linear (ablation)KIRI · MSCOCO (70% sparsity, LLaVA-1.5-7B)CIDEr: 78.2 -> 84.6Improved
use cubic sparsity schedule instead of constant (ablation)KIRI · MSCOCO (70% sparsity, LLaVA-1.5-7B)CIDEr: 76.5 -> 84.6Improved
add structural norms to DNA metric (ablation)KIRI · MSCOCO (70% sparsity, LLaVA-1.5-7B)CIDEr: 80.1 -> 84.6Improved
add activation term to DNA metric (ablation)KIRI · MSCOCO (70% sparsity, LLaVA-1.5-7B)CIDEr: 79.5 -> 84.6Improved
use DNA instead of magnitude-only importance (ablation)KIRI · MSCOCO (70% sparsity, LLaVA-1.5-7B)CIDEr: 77.3 -> 84.6Improved
add low-rank reconstruction (ablation)KIRI · MSCOCO (70% sparsity, LLaVA-1.5-7B)CIDEr: 81.3 -> 84.6Improved
add novelty reward to RL (ablation)KATANA · MSCOCO (70% sparsity, LLaVA-1.5-7B)CIDEr: 79.8 -> 84.6Improved
add diversity reward to RL (ablation)KATANA · MSCOCO (70% sparsity, LLaVA-1.5-7B)CIDEr: 80.4 -> 84.6Improved
add curriculum weighting to RL (ablation)KATANA · MSCOCO (70% sparsity, LLaVA-1.5-7B)CIDEr: 77.9 -> 84.6Improved
use GRPO policy updates instead of random evolutionary search (ablation)KATANA · MSCOCO (70% sparsity, LLaVA-1.5-7B)CIDEr: 76.2 -> 84.6Improved
reduce search from 1,000 to 100 iterationsKATANA · MSCOCO (70% sparsity, LLaVA-1.5-7B)CIDEr: 84.6 -> 78.2Hurt
reduce calibration set from N=128 to N=64 (ablation)KIRI · MSCOCO (70% sparsity, LLaVA-1.5-7B)CIDEr: 1.8 CIDEr dropHurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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