SemanticSpec introduces a semantic probability estimation mechanism that probes the model's internal hidden states to assess the likelihood of generating sequences with specific meanings, and proposes SemanticSpec, a semantic-aware speculative decoding framework that verifies entire semantic sequences instead of tokens...
Xi-Ming Dong, Shao-Wei Wang, Dayi Lin et al.· arXiv.org· 3 citations
The Spread--Novelty--Centrality (SNC) profile is introduced, a three-axis characterization of the demands of repository-level coding tasks, grounded in empirical software engineering research, and an unreliable proxy for task demands is introduced.
Radin Shayanfar, Keheliya Gallaba, Ahmed E. Hassan· 2 citations
MindForge is introduced, an automated pipeline that converts open-source command-line programs into source-free environments that expose only a compiled reference executable and its documentation that consistently improves over the base model across all seven unseen software engineering benchmarks, spanning long-horizo...
Yihao Chen, Shi Chang, Khaled Chawa et al.· arXiv.org· 0 citations
From an industrial code-generation improvement effort, a maintainer's perspective on why this work is hard in practice is offered, distilling three recurring challenges, zero-sum mixture design, yield as the binding metric, and end-to-end integration under uncertainty, and arguing that progress depends less on one-off...
Gopi Krishnan Rajbahadur, A. M. Ebrahimi, Boyuan Chen et al.· 0 citations
This work proposes LLMSafeGuard, a lightweight real-time framework that integrates an external validator into decoding, rejecting unsafe outputs while allowing valid ones, and introduces a similarity-based validation approach, simplifying safety constraint validation and eliminating the need for external control model...
Ximing Dong, Shaowei Wang, Dayi Lin et al.· SIGSOFT FSE Companion· 0 citations
Teaching deletion during post-training reduces deletion avoidance and improves broader code-editing performance, suggesting the behavior is undertrained rather than beyond reach.
A. M. Ebrahimi, M. M. Hasan, Aaditya Bhatia et al.· arXiv.org· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
