RT-SEG — Reasoning Trace Segmentation

RT-SEG is a modular research framework for the segmentation of reasoning traces into coherent structural and semantic units.

It provides:

  • Rule-based segmentation
  • Probabilistic and distributional boundary detection
  • LLM-driven discourse schemas
  • Topic and semantic shift segmentation
  • Late-fusion of heterogeneous engines
  • Reproducible database-backed experimentation

Design Philosophy

RT-SEG treats reasoning traces as structured discourse objects.
Segmentation hypotheses are represented as character-level offsets to guarantee:

  • deterministic reconstruction
  • reproducible evaluation
  • engine-agnostic comparison
  • systematic late fusion

Quickstart

from rt_seg import RTSeg, RTRuleRegex

segmentor = RTSeg(engines=RTRuleRegex)
offsets, labels = segmentor("Example reasoning trace...")

See the navigation sidebar for full documentation.


This site uses Just the Docs, a documentation theme for Jekyll.