Pith. sign in

REVIEW 20 cited by

lmgame-Bench: How Good are LLMs at Playing Games?

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2505.15146 v2 pith:OI6QZSLI submitted 2025-05-21 cs.AI

lmgame-Bench: How Good are LLMs at Playing Games?

classification cs.AI
keywords gameslmgame-benchllmsperceptioncontaminationevaluationgamememory
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Playing video games requires perception, memory, and planning, exactly the faculties modern large language model (LLM) agents are expected to master. We study the major challenges in using popular video games to evaluate modern LLMs and find that directly dropping LLMs into games cannot make an effective evaluation, for three reasons -- brittle vision perception, prompt sensitivity, and potential data contamination. We introduce lmgame-Bench to turn games into reliable evaluations. lmgame-Bench features a suite of platformer, puzzle, and narrative games delivered through a unified Gym-style API and paired with lightweight perception and memory scaffolds, and is designed to stabilize prompt variance and remove contamination. Across 13 leading models, we show lmgame-Bench is challenging while still separating models well. Correlation analysis shows that every game probes a unique blend of capabilities often tested in isolation elsewhere. More interestingly, performing reinforcement learning on a single game from lmgame-Bench transfers both to unseen games and to external planning tasks. Our evaluation code is available at https://github.com/lmgame-org/GamingAgent/lmgame-bench.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 20 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Beyond the Current Observation: Evaluating Multimodal Large Language Models in Controllable Non-Markov Games

    cs.CV 2026-06 unverdicted novelty 7.0

    RNG-Bench evaluates MLLMs on hidden-observation reconstruction in non-Markov games, finds forgetting as the dominant error source, and shows fine-tuning on optimal rollouts improves performance with transfer to other ...

  2. RTSGameBench: An RTS Benchmark for Strategic Reasoning by Vision-Language Models

    cs.AI 2026-06 unverdicted novelty 7.0

    RTSGameBench is a new extensible benchmark for VLMs using diverse RTS matchups, diagnostic mini-games targeting individual competencies, and a self-evolving query-to-game generator, with results showing poor VLM perfo...

  3. WebGameBench: Requirement-to-Application Evaluation for Coding Agents via Browser-Native Games

    cs.AI 2026-05 unverdicted novelty 7.0

    WebGameBench is a new benchmark that evaluates coding agents on building browser-native games from frozen specifications, with runtime browser evaluation showing best agents reach 76.9% usable rate but only 20.2% exce...

  4. WebGameBench: Requirement-to-Application Evaluation for Coding Agents via Browser-Native Games

    cs.AI 2026-05 unverdicted novelty 7.0

    WebGameBench is a benchmark that evaluates coding agents by having them generate browser-native games from specifications, then running those games in a real browser to assign EXCELLENT, USABLE, or UNUSABLE labels, wi...

  5. MMSkills: Towards Multimodal Skills for General Visual Agents

    cs.AI 2026-05 unverdicted novelty 7.0

    MMSkills packages multimodal procedural knowledge into state-conditioned skills with text, state cards, and multi-view keyframes, generated from public trajectories via an agentic process and used at inference via bra...

  6. MMSkills: Towards Multimodal Skills for General Visual Agents

    cs.AI 2026-05 unverdicted novelty 7.0

    MMSkills creates compact multimodal skill packages from trajectories and uses a branch-loaded agent to improve visual decision-making on GUI and game benchmarks.

  7. Co-Evolving LLM Decision and Skill Bank Agents for Long-Horizon Tasks

    cs.AI 2026-04 unverdicted novelty 7.0

    COSPLAY co-evolves an LLM decision agent with a skill bank agent to improve long-horizon game performance, reporting over 25.1% average reward gains versus frontier LLM baselines on single-player benchmarks.

  8. TriEx: A Game-based Tri-View Framework for Explaining Internal Reasoning in Multi-Agent LLMs

    cs.CL 2026-04 unverdicted novelty 7.0

    TriEx is a tri-view explainability framework that reveals systematic mismatches between what multi-agent LLMs say, believe, and do in imperfect-information strategic games.

  9. CAST: Game Solvers as Turn-Level Teachers for LLM Agents

    cs.CL 2026-07 conditional novelty 6.0

    CAST converts a game solver's per-action cost-to-go changes into turn-level RL credits for LLM agents and reports gains over outcome-only RLVR on three games plus zero-shot transfer to ALFWorld and WebShop.

  10. OmniGameArena: A Unified UE5 Benchmark for VLM Game Agents with Improvement Dynamics

    cs.CV 2026-06 unverdicted novelty 6.0

    OmniGameArena is a unified UE5 benchmark with 12 games and the IDC harness for cold-start scores and improvement dynamics of VLM agents.

  11. AgentOdyssey: Open-Ended Long-Horizon Text Game Generation for Test-Time Continual Learning Agents

    cs.CL 2026-05 unverdicted novelty 6.0

    Introduces AgentOdyssey, a procedural generator of open-ended long-horizon text games, to evaluate test-time continual learning agents and diagnose limits in exploration, memory, and planning.

  12. MINDGAMES: A Live Arena for Evaluating Social and Strategic Reasoning in Multi-Agent LLMs

    cs.AI 2026-05 unverdicted novelty 6.0

    Mindgames introduces a four-game evaluation platform for multi-agent LLM reasoning, runs a 944-agent competition, surfaces rule-adherence and error-survival limitations, and releases a 29k-game dataset with an offline...

  13. MMSkills: Towards Multimodal Skills for General Visual Agents

    cs.AI 2026-05 unverdicted novelty 6.0

    MMSkills turns public interaction trajectories into compact multimodal skill packages that visual agents can consult at runtime to improve decision-making on benchmarks.

  14. Odysseus: Scaling VLMs to 100+ Turn Decision-Making in Games via Reinforcement Learning

    cs.LG 2026-05 unverdicted novelty 6.0

    Odysseus adapts PPO with a turn-level critic and leverages pretrained VLM action priors to train agents achieving at least 3x average game progress over frontier models in long-horizon Super Mario Land.

  15. Agentic Environment Engineering for Large Language Models: A Survey of Environment Modeling, Synthesis, Evaluation, and Application

    cs.CL 2026-06 unverdicted novelty 5.0

    This survey categorizes agentic environments for LLMs by eight attributes and domains, introduces symbolic and neural synthesis paradigms with evaluation, and outlines four agent evolution pathways plus three environm...

  16. Robots Need More than VLA and World Models

    cs.RO 2026-06 unverdicted novelty 5.0

    The paper identifies four missing interfaces (data autolabelling, embodiment retargeting, physics-grounded world models, and video-based reward inference) as the central bottleneck beyond VLA scaling for robot intelligence.

  17. PTCG-Bench: Can LLM Agents Master Pok\'emon Trading Card Game?

    cs.AI 2026-05 unverdicted novelty 5.0

    PTCG-Bench shows LLM agents reach non-trivial PTCG performance but struggle with sustained self-evolution and remain sensitive to harness design.

  18. Gym-V: A Unified Vision Environment System for Agentic Vision Research

    cs.CV 2026-03 unverdicted novelty 5.0

    Gym-V supplies 179 visual environments showing that observation scaffolding like captions and rules matters more for training success than the choice of RL algorithm.

  19. UI-TARS-2 Technical Report: Advancing GUI Agent with Multi-Turn Reinforcement Learning

    cs.AI 2025-09 conditional novelty 5.0

    UI-TARS-2 reaches 88.2 on Online-Mind2Web, 47.5 on OSWorld, 50.6 on WindowsAgentArena, and 73.3 on AndroidWorld while attaining 59.8 mean normalized score on a 15-game suite through multi-turn RL and scalable data generation.

  20. A Survey of Reinforcement Learning for Large Reasoning Models

    cs.CL 2025-09 accept novelty 3.0

    A survey compiling RL methods, challenges, data resources, and applications for enhancing reasoning in large language models and large reasoning models since DeepSeek-R1.