大模型生成符号图形程序的新研究，直击自然语言到精准视觉内容的转化难题。
arXiv:2509.05208v2 Announce Type: replace-cross Abstract: Large language models (LLMs) excel at program synthesis, yet their ability to produce symbol…
实时编译器反馈让AI写代码更精准，降低迭代成本
arXiv:2607.13921v1 Announce Type: cross Abstract: Languages with rich static semantics, such as Rust, provide stronger guarantees for AI-generated cod…
用程序合成方法解释注意力机制，为理解Transformer内部行为提供新思路
arXiv:2606.19317v1 Announce Type: new Abstract: A longstanding goal of research on interpretable deep learning is to replace opaque neural computation…
LLM生成代码的视觉验证新思路：用TikZ案例探索不完美但有效的代码编辑验证方法
arXiv:2606.15693v1 Announce Type: cross Abstract: LLMs have significantly advanced code generation, enabling the synthesis of functional programs. Whi…
从输入输出行为自动合成程序，Discovery Agent挑战比NL2Code更难的IO2Code任务。
arXiv:2605.15334v1 Announce Type: cross Abstract: The automatic synthesis of a program from any form of specification is regarded as a holy grail of c…
提出属性引导方法，让LLM程序合成在规划任务中摆脱盲目分数依赖，提升搜索效率与成功率
arXiv:2605.16142v1 Announce Type: new Abstract: LLMs have shown impressive success in program synthesis, discovering programs that surpass prior solut…