AI
SeqFeed: Improving Agentic RTL Code Generation with Sequential Behavior Feedback
Researchers have proposed a new method called SeqFeed to improve the generation of RTL (Register-Tra...
AI
J-Miner: Recovering Executable Decision Knowledge from Language-Model Classifiers
Researchers have developed a method called J-Miner to extract decision-making knowledge from large l...
AI
Decomposition Attacks Across Unlinkable Identities: Limits of Stateful Defenses for LLM Services
Researchers have found a way to bypass defenses in large language model services by splitting harmfu...
AI
Code as Representation: A Compilable Parsing Paradigm for Academic Documents
Researchers have proposed a new approach to parsing academic documents called Compilable Academic Do...
AI
Reflex-Guard: A Low-Latency Guardrail for LLM Prompt Safety Using Dense Semantic Embeddings
Researchers have developed Reflex-Guard, a lightweight guardrail for Large Langu...
AI
An Empirical Study of Reward Specification and Benchmark Reliability in GRPO-based LLM Unlearning
Researchers have investigated the problem of specifying a reward function for la...
AI
Understanding Undesirable Word Embedding Associations
Researchers have been critical of word embeddings, which are used to represent w...
AI
An Analysis of Language Frequency and Error Correction for Esperanto
Researchers have analyzed the frequency of grammar errors in Esperanto using a d...