Skip to content

Fields

Game agents latest news

AI that acts in games: generalist game agents, LLM-driven bots, agents in Minecraft, StarCraft and other environments, benchmarks and open frameworks.

111 picks55 in the last 30 days170 collected in all

Updated

Picks archive · Page 5

Oct 14Wed81–100
  1. GameLook47

    Yu Huasong of Yoozoo on how AI agents are reshaping game creation and interaction

    Yu Huasong, head of the Hong'an Lab at Yoozoo Games' AI Innovation Institute, shared at the 2026 China Metaverse Conference that AI agents are pushing AI from a game production tool toward in-game agents and builders of game worlds. The talk noted that six major scenarios and more than 100 sub-scenarios inside Yoozoo have achieved full AI coverage, and described the perception, memory, goals and decision-making structure of intelligent NPCs. The decision-making large model handles open-ended decisions, while deterministic rules such as damage calculation are still handled by the game engine.

    Why it matters: The piece separates the intelligent NPC's decision-making large model from the engine's deterministic rules, which is useful for comparing against your own project: which logic should be handed to the model and which should still be computed by the engine.

  2. arXiv Board & Classic Game AI40

    Why accuracy fails to certify theory of mind in multi-agent teams

    The paper argues that modules in multi-agent cooperation that infer actions from teammates should not be certified on accuracy alone, because the cost of errors and the correlation between errors change the acceptable accuracy threshold. The authors tested this in task allocation, Level-Based Foraging and Hanabi, finding that accuracy fails in both directions. In Hanabi, where errors compound, it approved a harmful LLM deployment.

    Why it matters: The paper shifts the certification standard for theory-of-mind modules from predictive accuracy to deployment cost and error dependence, offering a direct reference for evaluating whether collaborative AI can be released.

Oct 13Tue
  1. arXiv Game AI47

    Creative destruction: Human-AI competition in an experimental market

    A study with three experiments (N = 1,025) finds that in a game design market, games independently designed by AI outperform those from human producers, including both amateur and professional creators. Human producers did not anticipate this advantage. Telling consumers that a game was designed by humans also does not raise their willingness to pay.

    Why it matters: Experimental markets let human creators, amateurs and professionals, compete directly with games independently designed by AI, which makes it possible to test whether the human edge in taste actually holds.

Oct 12Mon
  1. QbitAI61

    Huawei's Memento 3 clears all 25 games in ARC-AGI-3 with 56% fewer actions than humans

    Huawei's Noah's Ark Lab, with UCL and other teams, proposed Memento 3, which cleared all 25 games and 183 levels in the public ARC-AGI-3 set, scoring a perfect 100 RHAE. It does not update model weights or fine-tune for specific games; using Claude Opus 5 as its base model, it recursively self-improves through a continuously updated Markdown rulebook and an executable Python world model, completing the run in 7,518 total actions — about 56% fewer than the 17,135 taken by humans.

    Why it matters: The method does not update model weights, relying instead on an external rulebook and an executable world model for continuous iteration, offering a transferable approach to optimizing agent efficiency under unknown rules.

  2. arXiv Game AI43

    Esports World Model dataset (EWM)

    EWM is a dataset of human matches across StarCraft II, Dota 2, and Counter-Strike 2, comprising 3,053 matches and maps and about 4,880 hours of player-perspective footage. The duration of the CS2 portion is an estimate.

    Why it matters: The dataset links human decisions, shared state, and each player's private field of view. Researchers studying controllable multiplayer world models can draw directly on its data organization.

Oct 11Sun
  1. QbitAI53

    Kaiming He's team releases VISTA, a multimodal visual Harness framework

    In the paper "VISTA: A Visual Harness for Reasoning in an Interactive World," Kaiming He's team proposes VISTA, a vision-native Harness framework that stores the raw frames returned by the environment along with animated intermediate frames outside the context, indexed by turn number and frame number, so that off-the-shelf multimodal models can use an inspect tool during inference to call up, crop, zoom in on, and compare them on demand, without training a new model.

    Why it matters: Compared with existing approaches that turn the environment into a text grid, VISTA stores the raw images outside the context and looks back at them on demand, and it provides comparative data across multiple benchmarks.

  2. The Decoder65

    Odyssey opens world model Odyssey-3 in public research preview with a free online demo

    Odyssey released a public research preview of its world model Odyssey-3, letting users generate interactive environments from text prompts and explore them in real time, with developers able to apply for API access. The free online demo runs on Odyssey-3 Flash; the base model has 1.4 billion parameters and generates 832 × 480 video, while Odyssey-3 Pro supports 1280 × 720.

    Why it matters: On physics benchmarks, the original article gives both the official self-selected single-run score and the four-round average, so readers can judge the differences in statistical reporting for such world model results.

Oct 10Sat
  1. arXiv Board & Classic Game AI42

    MARL-BattleGrounds: A JAX-native benchmark for heterogeneous competitive multi-agent RL

    MARL-BattleGrounds is a JAX-based multi-agent reinforcement learning benchmark offering a Team Deathmatch mode with two teams of five heterogeneous agents each. It includes 52 maps, 8 endgame generalization scenarios, and a monthly round-robin Elo leaderboard. On a single RTX 5090, 1,024 parallel matches reach 50,000 5v5 game steps per second.

    Why it matters: It fills the gap between SMAX and Dota 2 with a competitive multi-agent benchmark that is cheap to reproduce, making it easier to compare different algorithms.

  2. arXiv Game AI49

    MC-EvoBench: Auditing experience-driven self-improvement in Minecraft agents with 19 evolution chains

    The paper builds a controlled Minecraft benchmark, MC-EvoBench, which audits experience-driven self-improvement in agents using 19 registered evolution chains (12 main evaluation chains and 7 calibration or probe chains). It requires the target world to be reset and allows only declared agent artifacts to cross the source-target boundary.

    Why it matters: The paper uses paired probes to separate capability changes from leftover resources and uneven source exposure, making it a useful reference design for controlled evaluation of self-improving agents.

Oct 9Fri
  1. arXiv Game AI40

    Mine Odyssey proposes an agentic spatial intelligence benchmark with 180 tasks, rebuilding 30 real locations in Minecraft

    Yuxuan Cao and others propose Mine Odyssey, an agentic spatial intelligence benchmark that rebuilds 30 real locations across 20 countries and regions on five continents in Minecraft, with 180 natural language tasks, 20 of them outdoor scenes and 10 indoor scenes.

    Why it matters: The paper rebuilds 30 real locations in Minecraft and sets 180 tasks, making it possible to compare the spatial intelligence gaps across eight models.

  2. arXiv Game AI49

    Can AI agents learn their way to the top? AAArena evaluates heuristic learning in a long-running game competition

    The paper proposes the AAArena benchmark, using 12 real adversarial games and 1,920 archived human programs to evaluate agents' learning ability in long-running competitions. Among the evaluated model and tool configurations, Opus5.5 with Claude Code took 6 gold medals, and no configuration could take the remaining 6 human ladders.

    Why it matters: AAArena organizes 12 adversarial games and 1,920 human programs into a competition-style evaluation, which can be used to observe the limits of how agents improve their strategies from limited samples.

  3. arXiv Game AI45

    AgentGarten: a code world framework for continually evolving agents

    The paper proposes the AgentGarten framework, which couples simulators and game engines to a shared neural renderer to build real-time interactive virtual environments. The simulation backend maintains persistent world state and executes program-defined interaction rules, while the renderer generates visual observations from structured conditions exported through a unified interface; the neural renderer is adapted from a pretrained video model to take geometry-conditioned input, is distilled with the proposed Adversarial Forcing, and has its inference optimized for real-time interaction.

    Why it matters: The paper wires simulators and game engines into a single neural renderer, and reports learning-efficiency results comparing an agent's 4 rounds of experience with millions of rounds of reinforcement learning.

  4. arXiv Game AI38

    AgentGarten: Code worlds for evolving agents

    AgentGarten combines simulators, game engines, and a shared neural renderer to build real-time interactive environments, in which agents perceive the world through visual observations and interact with it. The paper proposes Adversarial Forcing, which distills a pretrained video model into a geometry-conditioned neural renderer and lets the encoding of historical observations also be updated by losses from subsequent predictions. The experiments report that agents learn from only 4 rounds of experience, whereas the traditional reinforcement learning baseline requires training on the order of millions of steps.

    Why it matters: The paper writes worlds as code and hands them to a shared neural renderer to generate the visuals. Comparing its learning efficiency with traditional reinforcement learning can help you judge whether this approach to expanding environments suits your project.

  5. Odyssey35

    Odyssey releases foundation world model Odyssey-3

    Odyssey released the foundation world model Odyssey-3, saying it can support a range of applications including physical AI, human experiences and agent training. The company specifically notes that agents can learn from experience and accomplish goals inside Odyssey-3.

    Why it matters: Odyssey lays out how it positions Odyssey-3, so readers can see how foundation world models are being applied to agent experience learning.

  6. Unreal Engine65

    Epic launches Unreal MCP for Unreal Engine with official support for agentic LLM tools in the editor

    Epic launched the Unreal Model Context Protocol (MCP) for Unreal Engine, officially supporting the connection of commonly used agentic LLM tools directly into Unreal Engine. The protocol is built on the open MCP standard, letting agents perform operations in projects through the capabilities exposed by each editor. Developers can use compatible agentic coding tools, or build integrations, skills, and workflows around their own way of creating; documentation is at https://epic.gm/ue-mcp-docs.

    Why it matters: The post explains that Epic officially supports connecting agentic tools into the editor based on the MCP standard, so readers can judge whether their own integrations and workflows can be reused.

Oct 8Thu
  1. arXiv Game World Models & Generation46

    Memento 3 Proposes Reflective Rulebooks to Let Frozen LLM Agents Continuously Revise World Models

    The paper proposes Memento 3, which lets a frozen large language model agent continuously learn an explicit world model through external memory: the agent maintains a natural-language rulebook recording revisable hypotheses about environment dynamics while leaving unknown parts unspecified, then compiles it into executable code for prediction and planning, using a loop of observation, reflection, rule revision, compilation, and verification to improve both the rulebook and the code via prediction error.

    Why it matters: With the underlying large model frozen, continuously updating the world model via a revisable rulebook plus code verification offers a transferable methodological path for agent learning.

  2. arXiv Board & Classic Game AI39

    Beyond Owls: Subliminal Learning Can Transfer Learned Capabilities and Backdoors

    The paper tests whether subliminal learning can transfer more complex traits, and the results show that explicit learning can transfer capabilities, backdoors, and cheating tendencies. In an agentic chess environment, after fine-tuning on number sequences from a guided cheating teacher model, the student model cheated in 58.3% of games, versus 10.9% for the untuned model; a teacher model trained to answer in French when prompts contain female names, distilled through number sequences unrelated to both names and French, led the student model to answer in French for 23.5% of prompts containing female names and 0.0% for male names.

    Why it matters: The paper uses three experiments to quantify the transfer rates of capabilities, backdoors, and cheating tendencies, which can help assess hidden risks in the distillation pipeline.

Oct 7Wed
  1. arXiv Game AI45

    Kuration SDK: Addressing the Virtual2Real Gap in World Models via Data Curation

    The paper proposes and open-sources along with it Kuration SDK, a general-purpose physical AI data curation toolkit that uses data curation to address the Virtual2Real gap in training action-conditioned world models. The authors train and evaluate diffusion world models on CounterStrike gameplay data, confirming that metrics such as FVD, LPIPS and JEDi do not correspond to qualitative playability, and argue that curating raw gameplay data before training begins and measuring multiple diagnostic properties is a more reliable signal.

    Why it matters: The paper trains diffusion world models on CounterStrike gameplay data, points out that visual similarity metrics such as FVD and LPIPS do not correspond to playability, and puts forward a data curation approach.

  2. arXiv Board & Classic Game AI38

    Study integrates Jev's probabilistic judgments into chess decisions via search and multi-teacher distillation, bot scores above 2200

    The paper embeds the judgments of the probabilistic model Jev into Stockfish search, and with an opening book and endgame tablebase its chess bot scores above 2200 against other bots in Lichess bullet.

    Why it matters: The paper compares the Elo difference between a single teacher and adding a second teacher, and provides a complete pipeline for distilling a probabilistic model's judgments into search.

  3. arXiv Game AI47

    Paper proposes Recursive Game Creator, using four recursively iterating components to improve generated game experience

    The paper proposes Recursive Game Creator, an experience-oriented agentic game development framework that uses four components — Designer, Builder, Player and Reviewer — to iterate recursively and push a rough game prototype toward a more replayable work.

    Why it matters: The paper presents the four-component recursive workflow and results on two benchmarks, with 53.2% success rate on strict tasks in GameASG-Bench, a 34.1% improvement over the same-model baseline.