DigestAI news desk
hot OpenAI board member warns company is not on track to prevent catastrophic AI loss of control new R2VC Boosts Fact‑Checking Accuracy by 13.74% on FEVER Using Modular Retrieval and Calibration hot OpenAI Unveils GPT‑6 Astra: Record‑Breaking 3D Rendering, Loop‑Transformer Architecture new Study Finds Deictic Ambiguity Can Undermine Draft‑Verify‑Revise LLM Pipelines hot Mistral AI Secures $3.5B Series D, Valuation Tops $24B, Emphasizes Enterprise Control new New GLARE model improves meeting continuation forecasting on MDFB benchmark hot OpenAI delays IPO citing security risks after model escapes test environment new Chopthin-Consensus Power Sampling Boosts LLM Reasoning Accuracy Without Retraining hot OpenAI pledges to match Anthropic's embedded evaluator safety commitment new Occamy-1.0 Introduces Cost‑Efficient 35B Co‑Work Agent Model hot OpenAI launches Agents API in public beta, exposing Codex infrastructure to developers new Context-Augmented KG Training Boosts Multi-Hop QA Accuracy on Disease Graphs hot Meta launches Muse, a personal AI agent for everyday tasks on iOS and Android new Thermodynamical Genetic Algorithm Boosts LLM Artifact Efficiency by 29% on Bin‑Packing hot Anthropic reveals Houthis used Claude for ballistic missile development new HarnessDev Finds LLM‑Built Agent Harnesses Strong in Writing, Weak in Code Tasks
Research updated

Thermodynamical Genetic Algorithm Boosts LLM Artifact Efficiency by 29% on Bin‑Packing

Researchers introduce T‑GADE, a thermodynamical genetic algorithm that leverages large language model (LLM) operators to evolve structured artifacts—pairs of descriptions and code. By integrating free‑energy objectives and occupancy rules that control genotype repetition, the system balances diversity and generative power.

1 source primary source

Key points

  • T‑GADE combines thermodynamical genetic algorithms with LLM‑based operators to evolve structured description‑code artifacts.
  • On online bin‑packing, median training excess fell 29 % (1.152 % → 0.815 %) across 20 runs per configuration.
  • Statistical tests show significance (p = 0.042, Cliff’s δ = 0.378) and transfer excess matches baseline 0.496 %.

On the classic online bin‑packing benchmark, T‑GADE’s Bose‑type configuration at a temperature of 0.003 cut median training excess from 1.152 % to 0.815 %, a 29 % improvement over the Evolution of Heuristics (EoH) baseline. The study ran 20 independent trials per setting, yielding statistically significant gains (Mann‑Whitney p = 0.042, Cliff’s δ = 0.378). When the top two final candidates were validated on a transfer set with different bin capacities, the median excess matched EoH’s 0.496 %, confirming that thermodynamical selection can preserve performance while reducing training overhead.

Read the original at arXiv cs.AI · by Kyoko Ogawa, Naoki Mori primary source Open source ↗

The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us.

Comments

via GitHub Discussions

More in Research

All →

Related stories