LLMs
Latest breakthroughs in Large Language Models
Articles
Beyond Raw Transcripts: Structured Persona Extraction for LLM⭐9
Structured persona extraction improves LLM digital twin accuracy over raw transcripts; fixed schemas fail on diverse tasks but automatic discovery restores gain...
Automatic Recognition of Bioinformatics Software Names in⭐8
SNAIL framework accurately identifies bioinformatics software names in literature, outperforming existing tools and LLMs for scalable text mining.
Transformer Models for Text Summarization: A Comparative Study⭐9
Compare BERT, RoBERTa, and BART for extractive and abstractive text summarization, analyzing architectures and pretraining strategies.
Large Language Models in Mental Health: A Systematic Review of⭐9
This systematic review examines LLM applications in mental health, covering detection, therapy support, multimodal methods, and ethical deployment challenges.
A Virtual Member of a Community of Practice for the Society of⭐9
ATHENA, SPE’s AI assistant, boosts well-planning productivity. From prototype to deployment, learn how it enhances knowledge retrieval and collaboration.
LongNovel: A Multi-Scale Benchmark for Hallucination Detection⭐7
LongNovel benchmark tests hallucination detection in long-context novel summarization across Chinese and English, revealing how error patterns shift with contex...
Position: Collusion Risks Among AI Reasoning Agents Justify⭐7
AI reasoning agents risk tacit collusion in markets, eroding legal intent distinctions. This paper urges behavioral certification before deployment to prevent e...
Margin-Regularized Structured Semantic Alignment for Brain⭐7
MD-SigLIP aligns brain and text embeddings via margin-regularized structured ranking, improving retrieval-based brain-language decoding.
Auxiliary Uncertainty Signals for LLM-Assisted Systematic Review⭐9
Auxiliary BERT+GCN signals improve LLM screening in systematic reviews, with full-context delivery boosting F1 and MAYBE-only routing as the most cost-efficient...
FLOPs vs Real Work: The Importance of Replication in AI⭐10
FLOPs alone fail to predict AI model runtime. Replication reveals alpha-FLOPs underestimates modern hardware instability, urging transparent, reproducible effic...