CoderEval: A Benchmark of Pragmatic Code Generation with Generative Pre-trained Models
PulseAugur coverage of CoderEval: A Benchmark of Pragmatic Code Generation with Generative Pre-trained Models — every cluster mentioning CoderEval: A Benchmark of Pragmatic Code Generation with Generative Pre-trained Models across labs, papers, and developer communities, ranked by signal.
1 day(s) with sentiment data
-
New ACToR framework enhances AI code generation by targeting critical tokens
Researchers have developed a new framework called ACToR (Adaptive Critical Token-Aware Retrieval) to improve repository-level code generation. This method identifies critical tokens during the code generation process th…
-
Quantization impacts code generation models differently, study finds
A new study investigates the impact of various quantization methods on the performance of large code generation models when run on resource-constrained hardware. Researchers evaluated six state-of-the-art techniques, in…