Papers

Cards List

From constraint models to playable puzzle games

Lobsters Hottest · yesterday Cached

A researcher's blog post describes turning constraint models into playable puzzle games, based on his paper on scaling Sudoku as a constraint problem. It shares MiniZinc models, a repository of 434,201 Sudoku instances, and playable versions of nine puzzle games.

0 favorites 0 likes

Do Tabular Foundation Models Agree with Themselves?

arXiv cs.LG · yesterday Cached

This paper investigates whether tabular foundation models (TFMs) like TabPFN, TabICL, TabDPT, and TabFM produce predictions consistent with any joint distribution. It demonstrates that all evaluated TFMs violate both marginalization and factorization consistency for classification and regression, questioning their Bayesian inference claims.

0 favorites 0 likes

A Unified Risk View of Uncertainty: Posterior Risk for Disentanglement and Evaluation Beyond Proxies

arXiv cs.LG · yesterday Cached

This paper proposes a unified definition of uncertainty as pointwise posterior risk and introduces a theory-backed benchmark using semi-synthetic datasets to directly compute oracle epistemic and aleatoric uncertainty, enabling fine-grained evaluation beyond proxy tasks.

0 favorites 0 likes

Observation-Grounded Self-Predictive Reinforcement Learning for Visual Continuous Control

arXiv cs.LG · yesterday Cached

This paper proposes OG-SPR, a model-free visual RL algorithm that combines latent self-prediction with observation prediction to learn dynamics-aware representations, achieving improved sample efficiency on DeepMind Control Suite tasks.

0 favorites 0 likes

THBKG: A Temporal Biomedical Knowledge Graph for Decision-Aligned Clinical Advancement Prediction

arXiv cs.LG · yesterday Cached

This paper introduces THBKG, a temporal heterogeneous biomedical knowledge graph covering 110k entities and 11.1M edges with yearly evidence timestamps, designed to predict whether target-disease pairs entering Phase II trials advance to Phase III using only evidence available at the decision time. The graph-based approach outperforms direct-evidence baselines, especially for pairs lacking direct evidence, and the authors release it as a continually updated resource.

0 favorites 0 likes

How Far Do Simple Transformations Translate Across Text Embedding Models?

arXiv cs.LG · yesterday Cached

This arXiv paper investigates whether simple linear transformations can translate representations across nine heterogeneous text embedding models, finding that shared structure and transferability depend jointly on architecture, training objective, pooling, and data distribution, challenging the notion of universal latent compatibility.

0 favorites 0 likes

BioM-JEPA: joint-embedding prediction of graph-connected gene blocks in single cells

arXiv cs.LG · yesterday Cached

BioM-JEPA introduces a joint-embedding predictive architecture that learns single-cell representations by predicting graph-connected gene blocks instead of individual genes, showing improved efficiency and downstream performance in perturbation-response tasks.

0 favorites 0 likes

CohortHijack: Robustness of Single Cell Annotation to Companion Cell Removal

arXiv cs.LG · yesterday Cached

This arXiv paper introduces CohortHijack, a robustness audit that removes non-target cells from single-cell query cohorts to test how annotation tools can be manipulated without altering the target cell's expression profile. It shows that structured removal and search strategies can change refined labels in popular pipelines while preserving the target, identifying query cohort composition as a vulnerability surface.

0 favorites 0 likes

FormBharo: Designing and Evaluating a Voice Agent for Conversational Form Filling in Rural India

arXiv cs.CL · yesterday Cached

Presents FormBharo, a voice agent that uses LLMs with rule-based controls to fill structured forms over phone calls for low-literacy Hindi-speaking users in India, piloted with ARMMAN. The paper also introduces FormVoiceAgentBench, a benchmark of 3,760 multi-turn conversation tests, and shows that end-to-end evaluation is necessary since component-level performance does not predict full form completion.

0 favorites 0 likes

EpiBench: Can LLMs Understand Epitopes for Antibody Drug Discovery?

arXiv cs.CL · yesterday Cached

EpiBench is a new closed-book, sequence-based benchmark for evaluating how well LLMs understand epitopes across five antibody-drug-discovery tasks, finding that current models capture partial signals but struggle with antibody-specific reasoning.

0 favorites 0 likes

Clinical Communication Processing with Models Trained on LLM-Generated Synthetic Data: A Structured Survey and Novel Application Case Studies

arXiv cs.CL · yesterday Cached

This paper surveys clinical communication processing using LLM-generated synthetic data and presents 13 case studies across EMS reports, nurse handoffs, and more, showing that synthetic data can bootstrap clinical NLP systems.

0 favorites 0 likes

Alternating Levenberg-Marquardt Training of Physics-Informed Neural Networks with Fourier-Enhanced Features

arXiv cs.LG · yesterday Cached

This paper proposes FALM-PINN, an alternating Levenberg-Marquardt training framework for physics-informed neural networks that uses Fourier-enhanced features to address spectral bias and representation-coefficient coupling, achieving up to two orders of magnitude lower errors on high-frequency and nonlinear PDEs.

0 favorites 0 likes

Causal Episodic Memory for Feedback-Driven Agent Repair

arXiv cs.CL · yesterday Cached

This paper introduces MERIT, a training-free agent that uses causal episodic memory of past repair outcomes to improve subsequent Text-to-SQL generations, boosting execution accuracy on Spider and BIRD benchmarks.

0 favorites 0 likes

Beyond Feature Importance: A Comparative Analysis of Pattern Detection Methods in Cluster Interpretation

arXiv cs.LG · yesterday Cached

This paper presents a comparative evaluation of post-hoc analysis methods (Random Forest surrogate, LIME, PCA) for detecting structured patterns in clustering results, using synthetic datasets with injected patterns. It finds that none of the methods consistently detects all pattern types, highlighting a gap in existing explainability tools.

0 favorites 0 likes

MACRO: Markov Chain Routing of Transformer Layers

arXiv cs.CL · yesterday Cached

MACRO is a framework that learns task-specific execution routes over frozen LLM layers using Markov chain-based routing, improving reasoning accuracy without modifying model weights. It outperforms prior routing approaches while reducing search time significantly.

0 favorites 0 likes

Mapping Similarity Spaces across Embedding Models with Synthetic Query Probing

arXiv cs.CL · yesterday Cached

Introduces Synthetic Query Probing (SQP), a reference-free method for learning mappings between similarity score distributions across embedding models, enabling threshold portability in RAG systems. Experiments show systematic score distortions and isotonic regression as the best calibration approach.

0 favorites 0 likes

Evidential Rule Learning for Interpretable Classification with Abstention

arXiv cs.LG · yesterday Cached

Introduces Fast Evidential Rule Learning (FERL), a method for interpretable classification that produces evidential outputs and can abstain when uncertain, with theoretical stability guarantees and strong empirical results across tabular and concept-bottleneck benchmarks.

0 favorites 0 likes

Learning to Rank Tensor Network Contraction Plans for GPU-Accelerated Quantum Circuit Simulation

arXiv cs.LG · yesterday Cached

A research paper presenting a learning-to-rank framework for selecting efficient tensor-network contraction plans for GPU-accelerated quantum circuit simulation, using gradient-boosted rankers trained from GPU measurements.

0 favorites 0 likes

MameLoshnLM: Yiddish Language Model and Evaluation Benchmark

arXiv cs.CL · yesterday Cached

This paper introduces MameLoshnLM, the first open-source 8B-parameter Yiddish language model, along with the Oytser pretraining corpus and Kashes evaluation benchmark. It demonstrates that continued pretraining on high-quality Yiddish data outperforms general multilingual models, highlighting the value of dedicated low-resource language modeling.

0 favorites 0 likes

Predicting Task Difficulty Without Rollouts

arXiv cs.LG · yesterday Cached

This paper proposes predicting task difficulty for LLM agents without running expensive rollouts, studying the problem across 17 agentic benchmarks and showing that token-level entropy is a useful predictive signal.

0 favorites 0 likes
← Previous
Next →
← Back to home

Submit Feedback