arXiv subjects

Saman Dehghan

Publications and source records attributed to Saman Dehghan.

3 recordsLinked to original sources

arXiv2026

Online Monitoring and Corrective Steering of Programming Agents

Fixing GitHub issues in large-scale projects is a long-horizon task, especially when a fix requires changes across multiple locations or the issue description lacks the information needed to localize and repair it. As a result, agents traverse long trajectories that are prone to inefficiency and error: they drift away from their intended plan, repeat failed actions, or terminate without a working patch. This paper proposes LivePlan to monitor, detect, and correct such behavioral inefficiencies and drifts in real time. LivePlan decouples judging from advising: a deterministic, rule-based monitor examines general signals over the trajectory to detect issues without invoking an LLM, and only when an issue is detected does it consult an advisor LLM for a high-level, next-step correction. This design avoids the misleading re-planning and costly interventions of prior approaches. We implement LivePlan on top of SWE-agent and evaluate it using five LLMs (three as executor agents and two as advisors) across SWE-bench Verified and SWE-bench Pro. Compared to vanilla SWE-agent, LivePlan notably improves issue resolution rates, achieving consistent gains of up to 15.2% (average: 9.9%), while incurring only an additional cost of $0.08 per instance. The additional solutions concentrate on medium and hard instances. LivePlan consistently outperforms alternative approaches in resolution rate, with minimal regression on already successful runs and new successes on problems that no baseline solves.

cs.SE↗

arXiv2026

From Plan to Action: How Well Do Agents Follow the Plan?

Agents are commonly instructed to follow a task-specific plan for guidance. However, it is unknown to what extent agents actually follow instructed plans. Without such an analysis, determining the extent agents comply with a given plan, it is impossible to assess whether a solution was reached through correct strategic reasoning or through other means, e.g., data contamination or overfitting to a benchmark. This paper presents the first extensive, systematic analysis of plan compliance in programming agents, examining 21,120 trajectories from SWE-agent across four LLMs on SWE-bench Verified and SWE-bench Pro under eight plan variations. Without an explicit plan, agents fall back on internalized workflows during training, which are often incomplete, overfit, or inconsistently applied. Providing the standard plan improves issue resolution, and we observe that periodic plan reminders can mitigate plan violations and improve task success. A subpar plan hurts performance even more than no plan at all. Surprisingly, inserting additional task-relevant phases in the early stage can degrade performance, particularly when these phases do not align with the model's internal problem-solving strategy. These findings call for fine-tuning paradigms that teach models to follow instructed plans, rather than encoding task-specific plans in them, so that they reason and act adaptively, rather than memorizing workflows.

cs.SE↗

arXiv2025

Translating Large-Scale C Repositories to Idiomatic Rust

Existing C to Rust translation techniques fail to balance quality and scalability: transpilation-based approaches scale to large projects but produce code with poor safety, idiomaticity, and readability. In contrast, LLM-based techniques are prohibitively expensive due to their reliance on frontier models (without which they cannot reliably generate compilable translations), thus limiting scalability. This paper proposes Rustine, a fully automated pipeline for effective and efficient repository-level C to idiomatic safe Rust translation. Evaluating on a diverse set of 23 C programs, ranging from 27 to 13,200 lines of code, Rustine can generate fully compilable Rust code for all and achieve 87% functional equivalence (passing 1,063,099 assertions out of 1,221,192 in test suites with average function and line coverage of 74.7% and 72.2%). Compared to six prior repository-level C to Rust translation techniques, the translations by Rustine are overall safer (fewer raw pointers, pointer arithmetic, and unsafe constructs), more idiomatic (fewer Rust linter violations), and more readable. When the translations cannot pass all tests to fulfill functional equivalence, human developers were able to complete the task in 4.5 hours, on average, using Rustine as debugging support.

cs.SE↗