AI Paper Atlas

Research map / Benchmarks / ALFWorld

ALFWorld leaderboard

20 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on ALFWorld. Best success rate: 98 (NSI, ICML 2026).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

ALFWorld · success rate

20 papers, each with its best reported number for its own method; higher is better.

#success rate ↑MethodPaperTopic
198NSI
GPT-4o
Lifting Traces to Logic: Programmatic Skill Induction with Neuro-Symbolic Learning for Long-Horizon Agentic Tasks
ICML 2026 · Table 1
LLM agents
297.71SAMPO
Qwen3-8B (SFT)
ARLArena: A Unified Framework for Stable Agentic Reinforcement Learning
ICML 2026 · Table 7
RL from feedback
393Trajectory bootstrapping + database and exemplar curation
GPT-4o-mini
Self-Generated In-Context Examples Improve LLM Agents for Sequential Decision-Making Tasks
NeurIPS 2025 · Table: Average success rate of
Agents, code and math
492.5CoPO (full)
Qwen2.5-7B-Instruct
Think Fast and Slow: Step-Level Cognitive Depth Adaptation for LLM Agents
ICML 2026 · Table 2
Agents, code and math
592.41T2PO
Qwen3-8B-RFT
T$^2$PO: Uncertainty-Guided Exploration Control for Stable Multi-Turn Agentic Reinforcement Learning
ICML 2026 · Table 1
RL from feedback
691.79AdaPlanner
GPT-3
AdaPlanner: Adaptive Planning from Feedback with Language Models
NeurIPS 2023 · Table (ALFWorld)
LLM agents
787.1ALMA (transferred)
GPT-5-mini
Learning to Continually Learn via Meta-learning Agentic Memory Designs
NeurIPS 2026 · Table 1 Bottom
Online and replay-based
885.9Early experience: self-reflection (SR)
Llama-3.2-3B
Agent Learning via Early Experience
ICML 2026 · Table 1
Human-AI and goal-conditioned
984LAC (w/ finetuning)
Gemma-7B
Enhancing Decision-Making of Large Language Models via Actor-Critic
ICML 2025 · Appendix, Table: Ablation stud
Agents, code and math
1082.4DreamGym-S2R GRPO (5K real)
Qwen2.5-7B
Scaling Agent Learning via Experience Synthesis
ICLR 2026 · Table 1
Visual and representation RL
10 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.