AI Paper Atlas

Research map / Multi-agent systems and games / LLM agents

WALL-EVE: World Alignment with Rule Learning in Visual Environments

Siyu Zhou, Tianyi Zhou, Yijun Yang, Deheng Ye, Chengqi Zhang, Jing Jiang, Guodong Long

ECCV 2026 · Multi-agent systems and games › LLM agents · approach: Agentic benchmarks and orchestration · 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
add rules to the world model of the MPC agent (ablation)WALL-EVE MPC agent · Minecraft TechTreesuccess rate (%): 38 -> 69Improved
add rules to the agent only (ablation)WALL-EVE · Minecraft TechTreesuccess rate (%): 37 -> 61Improved
also add rules to the agent on top of rule-aligned world model (ablation)WALL-EVE · Minecraft TechTreesuccess rate (%): 69 -> 67Hurt
add LLM world model without rules to MPC (ablation)LLM agent · Minecraft TechTreesuccess rate (%): 37 -> 38No clear effect
add visual modality input (ablation)WALL-EVE · Minecraft TechTreesuccess rate (%): 59 -> 69Improved
add the world model (ablation)WALL-EVE · Minecraft TechTreesuccess rate (%): 61 -> 69Improved
add rule set pruning (ablation)WALL-EVE · Minecraft TechTreesuccess rate (%): 12 -> 69Improved
raise rule set limit from 2 to 4 (ablation)WALL-EVE · Minecraft TechTreesuccess rate (%): 52 -> 64Improved
remove rule cap (no limit, 6 rules) vs limit 4 (ablation)WALL-EVE · Minecraft TechTreesuccess rate (%): 64 -> 69Improved
use GPT-4o vision instead of fine-tuned Mask R-CNN for perception (ablation)WALL-EVE · ALFWorld (visual)success rate (%): 75 -> 64Hurt
use GPT-4o vision instead of fine-tuned Mask R-CNN for perception (ablation)RAFA · ALFWorld (visual)success rate (%): 72 -> 56Hurt
add learned rules to the LLM world model (ablation)LLM world model · ALFWorldsuccess/failure accuracy (%): 73 -> 94Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 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