Coding Agent with Executable World Models Solves 15 of 25 ARC-AGI-3 Benchmark Games
Researchers developed a coding-agent system that uses executable Python world models to tackle ARC-AGI-3 benchmark games, fully solving 15 of 25 public games with GPT-5.5 and 8 with GPT-4. The agent maintains and refines a world model, verifies it against observations, and plans actions without any game-specific hand-coded logic. The results offer preliminary evidence that verifier-driven world models are a viable approach for general AI reasoning benchmarks.
A team of researchers has published a preprint, accepted at AGI-2026, describing a coding-agent architecture designed to tackle ARC-AGI-3, a benchmark intended to test general reasoning in AI systems. The agent maintains an executable Python world model that it continuously verifies against prior observations and refactors toward simpler abstractions, approximating a minimum description length (MDL) simplicity bias. Crucially, the system uses no game-specific code, heuristics, or hidden solutions — the same agent and prompts are applied uniformly across all games. Using GPT-5.5 with high reasoning effort, the agent fully solved 15 of 25 public games and achieved a mean per-game Relative Highest Action Efficiency (RHAE) of 58.12%; with GPT-4, it solved 8 games with a mean RHAE of 41.29%. The researchers also conducted a security audit of the evaluation harness to close unintended information leakage channels that could artificially inflate scores. Performance on the private validation set has not yet been assessed, and the authors release full run artifacts and code publicly.
What's missing
The study has not yet been evaluated on the private ARC-AGI-3 validation set, so generalization beyond the 25 public games is unknown. The paper does not report statistical significance or variance across runs, making it unclear how sensitive results are to stochastic model behavior.
What different sources said
- arXiv cs.AICenter
Executable World Models for ARC-AGI-3 in the Era of Coding Agents
Related
Gut Bacteria Enzyme Found to Break Down Heat-Processed Food Compounds, Producing Novel Biogenic Amines
Researchers have discovered that an enzyme in common gut bacteria can degrade N-epsilon-carboxymethyllysine (CML), a compound formed during thermal food processing, producing previously unknown biogenic amines. The enzyme, ornithine decarboxylase SpeC from enterobacteria, acts on CML and related modified lysine derivatives through a low-level 'underground' catalytic activity. This finding suggests a previously unrecognized communication axis between thermally processed dietary compounds and gut microbial physiology, with potential implications for host health.
Full-Length Gene Sequencing Reveals Two Distinct Bacterial Communities in Black-Legged Ticks Expanding Into Canada
Researchers used Oxford Nanopore full-length 16S rRNA gene sequencing to characterize the microbiome of Ixodes scapularis black-legged ticks collected in Nova Scotia, Canada, distinguishing between tick-adapted bacteria and environmentally acquired bacteria. The study comes as I. scapularis — the primary vector of Lyme disease — is rapidly expanding northward into Canada due to climate change. The findings suggest that environmentally derived bacteria in tick microbiomes are not mere contamination, which has implications for how tick microbiome data is collected and interpreted across surveillance studies.
Study Identifies Metabolic Link Between Cell Envelope Stress and Biofilm Formation in Bacteria
Researchers have discovered that the metabolite acetyl-CoA directly inhibits enzymes that degrade the bacterial signaling molecule c-di-GMP, connecting cell envelope biosynthesis stress to biofilm formation in Pseudomonas aeruginosa. The study found that sub-inhibitory concentrations of antibiotics targeting early peptidoglycan biosynthesis — but not other antibiotic classes — elevate c-di-GMP levels by reducing phosphodiesterase activity, with acetyl-CoA competing for the enzyme active site. Because the relevant enzyme domain is broadly conserved across bacterial species, this checkpoint mechanism may be widespread and could have implications for understanding antibiotic-induced biofilm responses.