AI Paper Atlas

Research map / Large language models / Agents, code and math

Sponge Tool Attack: Stealthy Denial-of-Efficiency against Tool-Augmented Agentic Reasoning

Qi Li, Xinchao Wang

ICML 2026 · Large language models › Agents, code and math · approach: Long-horizon LLM agents · 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
remove history buffer (judge-only) (ablation)STA · Δ Cap Hits, Qwen2-VL-7BΔ Cap Hits (%): 26.54 -> 18.42Improved
remove judge feedback (buffer-only) (ablation)STA · Δ Cap Hits, Qwen2-VL-7BΔ Cap Hits (%): 26.54 -> 14.76Improved
remove history buffer (ablation)STA · |Similarity|, Qwen2-VL-7B|Similarity|: 0.70 -> 1.58Improved
remove judge feedback (ablation)STA · |Similarity|, Qwen2-VL-7B|Similarity|: 0.70 -> 1.52Improved
use MPNet-base-v2 instead of all-MiniLM-L6-v2 as reward model (ablation)STA · Δ Cap Hits, Qwen2-VL-7BΔ Cap Hits (%): 26.54 -> 26.10No clear effect
use Instructor-XL instead of all-MiniLM-L6-v2 as reward model (ablation)STA · Δ Cap Hits, Qwen2-VL-7BΔ Cap Hits (%): 26.54 -> 26.32No clear effect
shrink history buffer to per-probe k=2, global k=16 (ablation)STA · Δ Cap Hits, Qwen2-VL-7BΔ Cap Hits (%): 26.54 -> 24.87Hurt
enlarge history buffer to per-probe k=8, global k=32 (ablation)STA · Δ Cap Hits, Qwen2-VL-7BΔ Cap Hits (%): 26.54 -> 27.21No clear effect
raise tool-calling budget from 15 to 40STA · Δ Cap Hits, Qwen2-VL-7BΔ Cap Hits (%): 26.54 -> 24.23Mixed
raise tool budget from 15 to 40STA · Δ Steps, gpt-4o-miniΔ Steps: 2.13 -> 1.65Hurt
use stronger agentic framework OctoTools vs AutoGenSTA · Attack reward, Qwen2-VL-7Battack reward: 2.17 -> 2.99Improved
apply STA to OctoToolsOctoTools · Task accuracy, Qwen2-VL-7Baccuracy: 47.27 -> 45.58Hurt

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