Context Pruning for Coding Agents via Multi-Rubric Latent Reasoning

TL;DR

This paper introduces LaMR, a structured pruning framework that decomposes code relevance into semantic and dependency dimensions, improving token efficiency and accuracy in coding agents by filtering irrelevant code segments.

Contribution

LaMR is a novel multi-rubric structured pruning method that models heterogeneous relevance patterns with dedicated CRFs and uses AST-based analysis for supervision, outperforming unpruned baselines.

Findings

LaMR matches or outperforms full-context baselines on four benchmarks.

It saves up to 31% tokens in multi-turn tasks.

It improves Exact Match scores by up to +3.5 on single-turn tasks.

Abstract

LLM-powered coding agents spend the majority of their token budget reading repository files, yet much of the retrieved code is irrelevant to the task at hand. Existing learned pruners compress this context with a single-objective sequence labeler, collapsing all facets of code relevance into one score and one transition matrix. We show that this formulation creates a modeling bottleneck: a single CRF transition prior must serve heterogeneous retention patterns, including contiguous semantic spans and sparse structural support lines. We propose LaMR (Latent Multi-Rubric), a structured pruning framework that decomposes code relevance into two interpretable quality dimensions, semantic evidence and dependency support, each modeled by a dedicated CRF with dimension-specific transition dynamics. A mixture-of-experts gating network dynamically weights the per-rubric emissions conditioned on the query, and a final CRF layer on the fused emissions produces the aggregate keep-or-prune decision. To supervise each dimension without additional annotation cost, we derive multi-rubric labels from the existing training corpus via AST-based program analysis, simultaneously denoising the teacher's binary labels. By effectively filtering distracting noise, LaMR frequently matches or even outperforms unpruned full-context baselines. Experiments on four benchmarks (SWE-Bench Verified, SWE-QA, LCC, LongCodeQA) show that LaMR wins 12 of 16 head-to-head multi-turn comparisons. It saves up to 31% more tokens on multi-turn agent tasks and improves Exact Match by up to +3.5 on single-turn tasks, while performance is frequently enhanced by denoising the context, and any remaining drops are marginal.