New Links | ML News

ML Newsnew | past | best | rss | submit

2041.	Token Embeddings Violate the Manifold Hypothesis (arxiv.org)
	3 points by lukeplato 329 days ago \| hide \| past \| pdf \| 1 comment
2042.	Gradient-Sign Masking for Task Vector Transport Across Pre-Trained Models (arxiv.org)
	2 points by PaulHoule 329 days ago \| hide \| past \| pdf \| discuss
2043.	TeaRAG: A Token-Efficient Agentic Retrieval-Augmented Generation Framework (arxiv.org)
	2 points by saikatsg 329 days ago \| hide \| past \| pdf \| 1 comment
2044.	Diagnosing Representation Dynamics in NER Model Extension (arxiv.org)
	1 point by PaulHoule 329 days ago \| hide \| past \| pdf \| discuss
2045.	Latent Topic Synthesis: Leveraging LLMs for Electoral Ad Analysis (arxiv.org)
	1 point by PaulHoule 329 days ago \| hide \| past \| pdf \| discuss
2046.	Automating Benchmark Design (arxiv.org)
	1 point by bbzjk7 330 days ago \| hide \| past \| pdf \| discuss
2047.	FPI-Det: A Face–Phone Interaction Dataset for Phone-Use Detection (arxiv.org)
	1 point by 0xK1K3 330 days ago \| hide \| past \| pdf \| discuss
2048.	Long-Context Attention from Kernel Efficiency to Distributed Context Parallelism (arxiv.org)
	1 point by PaulHoule 330 days ago \| hide \| past \| pdf \| discuss
2049.	The Principles of Diffusion Models (arxiv.org)
	239 points by Anon84 330 days ago \| hide \| past \| pdf \| 31 comments
2050.	Growing Cosine Unit (GCU): A Novel Oscillatory Activation Function (2023) (arxiv.org)
	1 point by peter_d_sherman 330 days ago \| hide \| past \| pdf \| discuss
2051.	Mathematical exploration and discovery at scale – Terence Tao et al. (arxiv.org)
	4 points by crvdgc 331 days ago \| hide \| past \| pdf \| 1 comment
2052.	LLMs Position Themselves as More Rational Than Humans (arxiv.org)
	2 points by ascorbic 331 days ago \| hide \| past \| pdf \| discuss
2053.	It Is All about Token: Towards Semantic Information Theory for LLMs (arxiv.org)
	3 points by dboreham 332 days ago \| hide \| past \| pdf \| discuss
2054.	Computational Turing test shows systematic difference between human, AI language (arxiv.org)
	2 points by anigbrowl 332 days ago \| hide \| past \| pdf \| discuss
2055.	Perplexitys First Research Paper – Point-to-Point Communication for LLM Systems (arxiv.org)
	1 point by Alifatisk 332 days ago \| hide \| past \| pdf \| discuss
2056.	Benchmarking World-Model Learning (arxiv.org)
	1 point by mpweiher 332 days ago \| hide \| past \| pdf \| discuss
2057.	Arima_plus: Better Time Series Forecasting and Anomaly Detection in BigQuery (arxiv.org)
	2 points by zeptonaut22 332 days ago \| hide \| past \| pdf \| discuss
2058.	From Memorization to Reasoning in the Spectrum of Loss Curvature (arxiv.org)
	65 points by andy12_ 332 days ago \| hide \| past \| pdf \| 14 comments
2059.	Valid Inference with Imperfect Synthetic Data (arxiv.org)
	1 point by djhu9 333 days ago \| hide \| past \| pdf \| discuss
2060.	Mathematical Exploration and Discovery at Scale (arxiv.org)
	4 points by tzury 333 days ago \| hide \| past \| pdf \| 2 comments
2061.	Diffusion Language Models Are Super Data Learners (arxiv.org)
	3 points by jonbaer 333 days ago \| hide \| past \| pdf \| discuss
2062.	Death by a Thousand Prompts: Open Model Vulnerability Analysis (arxiv.org)
	2 points by hsanthan 333 days ago \| hide \| past \| pdf \| 1 comment
2063.	Evaluating Control Protocols for Untrusted AI Agents (arxiv.org)
	1 point by timini 333 days ago \| hide \| past \| pdf \| 1 comment
2064.	HaluMem: Evaluating Hallucinations in Memory Systems of Agents (arxiv.org)
	2 points by timini 333 days ago \| hide \| past \| pdf \| 1 comment
2065.	LLMs encode how difficult problems are (arxiv.org)
	174 points by stansApprentice 333 days ago \| hide \| past \| pdf \| 38 comments
2066.	Mathematical Exploration and Discovery at Scale (arxiv.org)
	2 points by SweetSoftPillow 333 days ago \| hide \| past \| pdf \| discuss
2067.	Whisper Leak: a side-channel attack on Large Language Models (arxiv.org)
	3 points by bikenaga 333 days ago \| hide \| past \| pdf \| discuss
2068.	The Curved Spacetime of Transformer Architectures (arxiv.org)
	2 points by luis_likes_math 333 days ago \| hide \| past \| pdf \| 1 comment
2069.	Sequential Scaling Beats Majority Voting at Matched Compute (arxiv.org)
	2 points by paraschopra 333 days ago \| hide \| past \| pdf \| discuss
2070.	The OpenHands Software Agent SDK: Composable and Extensible (arxiv.org)
	1 point by timini 333 days ago \| hide \| past \| pdf \| 1 comment
	More