Tootfinder

Reasoning about Uncertainty: Do Reasoning Models Know When They Don't Know?
Zhiting Mei, Christina Zhang, Tenny Yin, Justin Lidard, Ola Shorinwa, Anirudha Majumdar
https://arxiv.org/abs/2506.18183

Reasoning about Uncertainty: Do Reasoning Models Know When They Don't Know?
Reasoning language models have set state-of-the-art (SOTA) records on many challenging benchmarks, enabled by multi-step reasoning induced using reinforcement learning. However, like previous language models, reasoning models are prone to generating confident, plausible responses that are incorrect (hallucinations). Knowing when and how much to trust these models is critical to the safe deployment of reasoning models in real-world applications. To this end, we explore uncertainty quantification…

Vikes' Metellus practices following spring 'hold-in' https://www.espn.com/nfl/story/_/id/45807863/josh-metellus-practices-vikings-spring-hold-in

Rethinking Click Models in Light of Carousel Interfaces: Theory-Based Categorization and Design of Click Models
Jingwei Kang, Maarten de Rijke, Santiago de Leon-Martinez, Harrie Oosterhuis
https://arxiv.org/abs/2506.18548

Rethinking Click Models in Light of Carousel Interfaces: Theory-Based Categorization and Design of Click Models
Click models are a well-established for modeling user interactions with web interfaces. Previous work has mainly focused on traditional single-list web search settings; this includes existing surveys that introduced categorizations based on the first generation of probabilistic graphical model (PGM) click models that have become standard. However, these categorizations have become outdated, as their conceptualizations are unable to meaningfully compare PGM with neural network (NN) click models …

Generalized Verma modules over sl(m 1) induced from simple highest weight modules
Yaohui Xue, Yan Wang
https://arxiv.org/abs/2506.19625 https://

Generalized Verma modules over sl(m+1) induced from simple highest weight modules
A class of generalized Verma modules over sl(m+1) are constructed from simple highest weight gl(m)-modules. Furthermore, the simplicity criterion for these sl(m+1)-modules are determined and an equivalence between generalized Verma modules and tensor modules are established.

Risk In Context: Benchmarking Privacy Leakage of Foundation Models in Synthetic Tabular Data Generation
Jessup Byun, Xiaofeng Lin, Joshua Ward, Guang Cheng
https://arxiv.org/abs/2507.17066

Risk In Context: Benchmarking Privacy Leakage of Foundation Models in Synthetic Tabular Data Generation
Synthetic tabular data is essential for machine learning workflows, especially for expanding small or imbalanced datasets and enabling privacy-preserving data sharing. However, state-of-the-art generative models (GANs, VAEs, diffusion models) rely on large datasets with thousands of examples. In low-data settings, often the primary motivation for synthetic data, these models can overfit, leak sensitive records, and require frequent retraining. Recent work uses large pre-trained transformers to …

Checklists Are Better Than Reward Models For Aligning Language Models
Vijay Viswanathan, Yanchao Sun, Shuang Ma, Xiang Kong, Meng Cao, Graham Neubig, Tongshuang Wu
https://arxiv.org/abs/2507.18624

Checklists Are Better Than Reward Models For Aligning Language Models
Language models must be adapted to understand and follow user instructions. Reinforcement learning is widely used to facilitate this -- typically using fixed criteria such as "helpfulness" and "harmfulness". In our work, we instead propose using flexible, instruction-specific criteria as a means of broadening the impact that reinforcement learning can have in eliciting instruction following. We propose "Reinforcement Learning from Checklist Feedback" (RLCF). From instructions, we extract checkl…

🇺🇦 #NowPlaying on KEXP's #MorningShow
Metallica:
🎵 Fade to Black
#Metallica
https://shonsta.bandcamp.com/track/metallica-fade-to-black-guitar-cover
https://open.spotify.com/track/5nekfiTN45vlxG0eNJQQye

Studie: Verändert KI die menschliche Sprache?
Binnen kurzer Zeit habe der Gebrauch von Large Language Models den Sprachgebrauch der Menschen verändert, warnen Forscher. Die Folgen könnten tiefgreifend sein.
https://…

Bayesian Social Deduction with Graph-Informed Language Models
Shahab Rahimirad, Guven Gergerli, Lucia Romero, Angela Qian, Matthew Lyle Olson, Simon Stepputtis, Joseph Campbell
https://arxiv.org/abs/2506.17788

Bayesian Social Deduction with Graph-Informed Language Models
Social reasoning - inferring unobservable beliefs and intentions from partial observations of other agents - remains a challenging task for large language models (LLMs). We evaluate the limits of current reasoning language models in the social deduction game Avalon and find that while the largest models demonstrate strong performance, they require extensive test-time inference and degrade sharply when distilled to smaller, real-time-capable variants. To address this, we introduce a hybrid reaso…

CLEAR-3K: Assessing Causal Explanatory Capabilities in Language Models
Naiming Liu, Richard Baraniuk, Shashank Sonkar
https://arxiv.org/abs/2506.17180 http…

CLEAR-3K: Assessing Causal Explanatory Capabilities in Language Models
We introduce CLEAR-3K, a dataset of 3,000 assertion-reasoning questions designed to evaluate whether language models can determine if one statement causally explains another. Each question present an assertion-reason pair and challenge language models to distinguish between semantic relatedness and genuine causal explanatory relationships. Through comprehensive evaluation of 21 state-of-the-art language models (ranging from 0.5B to 72B parameters), we identify two fundamental findings. First, l…