AI Wiki · updated daily
New AI research today
707 new papers were announced on arXiv today in AI, language and machine learning. Here are the first ten in each field, each linked to the paper. As of 2026-11-03.
Artificial intelligence
- A Generalized Sparse Fuzzy Clustering Framework with Automated Cluster Cardinality Estimation Ability
- Learning to bid and bundle: a posterior-sampling agent for online retail
- Early Stopping in Large Reasoning Models: A Survey and Empirical Audit of Efficiency and Reliability
- PatchScout: Thematic Web Data Collection via Information Foraging
- Reimagining Text-to-SQL: A Survey of Large Language Model Architectures and Strategies
- AuroraWorld: Structured 4D World Modeling for Exo-to-Ego Video Generation in Robot Manipulation
- Neurons Are All You Need
- MemU-Bench: Benchmarking GUI Agents on Long-Horizon User Interactions
- From Cross-Section to Cross-Omics: Conditional Cell Nucleus Image Editing in Spatial Transcriptomics
- RE-SciVQA: Improving Scientific VQA With Context-Relevant Retrieval and Lightweight MLLM Adaptation
Language
- From Fragmentation to Fusion: A Survey and Roadmap for Omnimodal Embeddings
- MEMENTO: Learning Context Management for Long Context Processing Agents
- Cross-Lingual Persona Effects in LLM Negotiation
- GAVAGAI: Referential Alignment for Infant-Scale Vision-Language Models via Balanced Optimal Transport
- Queering the Audits: Community-Based Auditing of AI Harms to Queer Communities
- Cross-Prompt Reuse of Entity-Replacement Responses
- Humanized but Still True? Auditing Facts and Detection in AI-Text Humanizers
- Circuit Oracle: Automating Attribution Graph Analysis via Natural-Language Queries
- Fidelity Collapses with Table Size, Not Turn Count: Delivery Architecture in a Deployed Tool-Using Agent
- From Lookups to Lessons: Turning Retrieved Knowledge into Parametric Memory
Machine learning
- Output-Aware Block Influence: Jacobian-Lens Weighting for Depth Pruning
- FB-MOAC: A Reinforcement Learning Algorithm for Forward-Backward Markov Decision Processes
- Interpretable structural temperature scaling for calibrating graph neural networks and point-cloud segmentation
- ProLand: A New Infeasible Manifold Optimization Method for LoRA Fine-Tuning
- IsoAlign: Isoform-Aware DNA-Protein Alignment via Joint-Embedding Predictive Architectures
- KINDER: Kernel-based Independence for Fair Representation Learning via Prototype-space Erasure
- DELBERT-2: Pretrained Fingerprint Language Models for DEL Protein Binder Prediction
- OCELOT: Planetary-Scale Species Distribution Modeling with Satellite Foundation Embeddings
- Quantization of Large Models with Weight Outliers and Sparsity
- A Framework for Task-Agnostic Counterfactual Fairness in Transformers through Activation Steering
Source: arXiv daily announcements, read 2026-11-03. Papers on arXiv are preprints and not yet peer-reviewed.