arXiv Machine Learning June 9, 2026 By Jasper Zhang, Bryan Cheng

Information-Theoretic Requirements for Gradient-Based Task Affinity Estimation in Multi-Task Learning

Read the original on arXiv Machine Learning →

arXiv:2604. 07848v2 Announce Type: replace Abstract: Multi-task learning shows strikingly inconsistent results -- sometimes joint training helps substantially, sometimes it actively harms performance -- yet the field lacks a principled framework for predicting these outcomes.

Summary generated by The Flow from the publisher's feed. The full article lives at arXiv Machine Learning.

arXiv AI

Jul 21

Capacity and Redundancy Trade-offs in Multi-Task Learning

arXiv:2607. 16554v1 Announce Type: cross Abstract: In multi-task learning (MTL) negative transfer is often considered as an optimization artifact, but it can also be viewed as a consequence of limited shared capacity and weak task redundancy.

By Asif Khan

fine-tuning

arXiv Machine Learning

Jul 2

Deep Multitask Learning for Mixed-Type Outcomes with Shared Sparsity

arXiv:2607. 00995v1 Announce Type: cross Abstract: Most existing multitask learning approaches are limited by their reliance on task-specific loss functions tailored to the scale and type of each outcome.

By Huichao Li, Tong Wang, Sanguo Zhang, Shuangge Ma

efficiency

arXiv AI

Jul 7

SNR-Adaptive Unified Diffusion for Multi-Task Medical Image Segmentation

arXiv:2607. 03103v1 Announce Type: cross Abstract: Clinical cardiac imaging pipelines currently deploy separate models for each dataset and modality, incurring redundant training costs and precluding knowledge sharing across anatomically related tasks.

By Jiahao Liu, Hang Wei, Shuai Wu

diffusion computer-vision fine-tuning benchmarks safety

arXiv Machine Learning

Jun 2

CORE-MTL: Rethinking Gradient Balancing via Causal Orthogonal Representations

arXiv:2606. 02221v1 Announce Type: cross Abstract: Multi-task learning (MTL) aims to construct a joint model for multiple tasks by sharing a common representation across domains.

By Chengfeng Wu, Tao Zou, Yanru Wu, Jingge Wang

benchmarks

arXiv Machine Learning

Jun 11

Probabilistic Contrastive Pretraining for Multi-task ADME Property Prediction

arXiv:2606. 11508v1 Announce Type: new Abstract: Accurate prediction of absorption, distribution, metabolism, and excretion (ADME) properties is critical to drug discovery, but remains challenging because ADME endpoints are noisy, interdependent, and often data-limited.

By Yifan Xue, Srimukh Prasad Veccham, Saee Paliwal, Tyler Shimko, Micha Livne

llms fine-tuning

arXiv Machine Learning

Jun 26

The Geometry of Updates: Fisher Alignment at Vocabulary Scale

arXiv:2606. 27242v1 Announce Type: new Abstract: Training-free source selection for LLM families with shared vocabularies arises in scientific string domains such as SMILES, protein, and genomic sequences, where candidate corpora share a tokenizer but differ in prediction targets.

By John Sweeney

llms rag nlp safety

Related stories

Capacity and Redundancy Trade-offs in Multi-Task Learning

Deep Multitask Learning for Mixed-Type Outcomes with Shared Sparsity

SNR-Adaptive Unified Diffusion for Multi-Task Medical Image Segmentation

CORE-MTL: Rethinking Gradient Balancing via Causal Orthogonal Representations

Probabilistic Contrastive Pretraining for Multi-task ADME Property Prediction

The Geometry of Updates: Fisher Alignment at Vocabulary Scale