paper-with-me

Papers

CRANE: Constrained Reasoning Injection for Code Agents via Nullspace Editing

2026-05-13 · Mingzhi Zhu, Michele Merler, Raju Pavuluri, Stacy Patterson arxiv

Code agents must both reason over long-horizon repository state and obey strict tool-use protocols. In paired Instruct/Thinking checkpoints, these capabilities are complementary but misaligned. The Instruct model is concise and tool-disciplined, whereas the Thinking model offers stronger planning and recovery behavior but often over-deliberates and degrades agent performance. We present CRANE (Constrained Reasoning Injection for Code Agents via Nullspace Editing), a training-free parameter-editing method that treats the Thinking-Instruct delta as a directional pool of candidate reasoning edits for the Instruct backbone. CRANE combines magnitude thresholding to denoise the delta, a Conservative Taylor Gate to retain edits that are jointly beneficial for reasoning transfer and tool-use preservation, and Graduated Sigmoidal Projection to suppress format-critical update directions. By merging paired Instruct and Thinking checkpoints, CRANE delivers strong gains over either individual model while preserving Instruct-level efficiency: on Roo-Eval it achieves pass1 of 66.2% (+19.5%) for Qwen3-30B-A3B and 81.5% (+8.7%) for Qwen3-Next-80B-A3B; on SWE-bench-Verified it resolves up to 14 additional instances at both scales (122/500 and 180/500); and on Terminal-Bench v2 it improves pass1/pass5 by up to 2.3%/7.8%, reaching 7.6%/17.9% and 14.8%/30.3%, respectively, consistently outperforming alternative merging strategies across all three benchmarks.

📄 PDF Abstract BibTeX arXiv:2605.14084

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

CRANE: Reasoning with constrained LLM generation

2025-02-13 · Debangshu Banerjee, Tarun Suresh, Shubham Ugare, Sasa Misailovic 외

Code generation, symbolic math reasoning, and other tasks require LLMs to produce outputs that are both syntactically and semantically correct. Constrained LLM generation is a promising direction to enforce adherence to …

Code GenerationMathvalid

CRANE: Knowledge Editing for Reasoning MLLMs

2026-06-08 · Han Huang, Hao Wang, Mengqi Zhang, Shu Wu 외 arxiv

The emergence of reasoning multimodal large language models (MLLMs), which generate explicit chain-of-thought (CoT) reasoning before producing answers, has introduced a new challenge for knowledge editing: methods that a…

knowledge editing

Modelling and Control of a Knuckle Boom Crane

2021-03-30 · M. Ambrosino, E. Garone

Cranes come in various sizes and designs to perform different tasks. Depending on their dynamic properties, they can be classified as gantry cranes and rotary cranes. In this paper we will focus on the so called 'knuckle…

Oscillation Reduction for Knuckle Cranes

2021-03-03 · Michele Ambrosino, Brent Thierens, Arnaud Dawans, Emanuele Garone

Boom cranes are among the most common material handling systems due to their simple design. Some boom cranes also have an auxiliary jib connected to the boom with a flexible joint to enhance the maneuverability and incre…

AWE: Adaptive Agents for Dynamic Web Penetration Testing

2026-03-01 · Akshat Singh Jaswal, Ashish Baghel arxiv

Modern web applications are increasingly produced through AI-assisted development and rapid no-code deployment pipelines, widening the gap between accelerating software velocity and the limited adaptability of existing s…