AI Paper Atlas

Research map / Large language models / Agents, code and math

Automated Repair of Totally-Ordered Hierarchical Task Network Domains via Context-Free Grammars with Large Language Model Support

Daniel Lutalo, Pascal Bercher

AAAI 2026 · Large language models › Agents, code and math · approach: Long-horizon LLM agents

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace symbolic re-encoding with LLM-based CFG repair (o4-mini)TO-HTN domain repair · TO-HTN repair benchmark (156 instances)solved instances: 76 -> 145Improved
use open-weight gpt-oss-120b instead of o4-miniLLM-based repair · TO-HTN repair benchmarksolved instances: 145 -> 140No clear effect
replace optimal encoding with suboptimal greedy encodingsymbolic repair encoding · TO-HTN repair benchmarksolved instances: 57 -> 76Improved
use LLM repair instead of suboptimal encoding on Monroe (FO)TO-HTN domain repair · Monroe (FO)solved instances: 0 -> 17Improved
use LLM repair instead of suboptimal encoding on ChildsnackTO-HTN domain repair · Childsnacksolved instances: 10 -> 9Hurt
use LLM repair instead of optimal encoding in terms of repair costTO-HTN domain repair · 57 instances with known optimal costrepair cost: optimal in 42 of 57 cases; 6% higher average costMixed
use greedy suboptimal encoding instead of optimal encoding for repair costsymbolic repair encoding · 57 instances with known optimal costrepair cost: 44% higher average costHurt
increase plan length (plans longer than about 40 steps)LLM-based repair · TO-HTN repair benchmarkfailure rate: plan length p = 0.015; token count p = 0.052Hurt
add structural filtering and CFG pruning before prompting the LLMLLM-based repair · TO-HTN repair benchmarkprompt size and coverageImproved
mask all action and method names in the promptLLM-based repair · TO-HTN repair benchmarkdata leakage riskImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy