AI

Topology-Consistent Task Planning over Cellular Workflow Complexes for LLM-based Agents

Read full article at source ↗

Based on a feed excerpt; full article content has not been acquired. Source evidence retrieved 08/10/2026, 02:18:24 UTC.

Briefings

This article introduces TopoPlanner, a framework for LLM-based agents that uses cellular workflow complexes to handle complex tool dependencies. Experiments on four benchmarks show consistent improvements over prompt-based and graph-enhanced baselines across different local LLM backbones.

Key points

• TopoPlanner lifts tool dependency graphs into cellular workflow complexes to serve as topology-aware context for LLM tool planning.
• The framework retrieves a request-relevant closed subcomplex via cosheaf-consistent cellular retrieval and performs multidimensional structural reasoning over the retrieved topology.

Context

Existing planners struggle with workflow patterns like verification-correction loops and convergent branch merging. TopoPlanner addresses this by using cellular workflow complexes as topology-aware context for LLM tool planning.

Limitations

Results are limited to four tool-planning benchmarks; improvements are shown across different local LLM backbones but specific performance metrics per backbone are not detailed.

Generated by Bittrees-hosted qwen35-2b sha256:aaf42c8b7c3c on 26/10/2026, 04:03:55 UTC. Based on publisher-provided feed evidence; may not cover the full original article. Check the source for details.

View archived briefing on IPFS
bafybeicvzaxp3kbctytefee2kgvrkqpktbehcibzxchtun36txcmdsulkm