Apple-Machine Learning Research Apple

1

Plan or Retrieve First? A Scale Study of Tool Discovery Architectures

2

Selection-Based Structured Reasoning: Toward Efficient Multimodal Search Agents

3

Code-Switching Detection via Language-Selective Dimensions in Speech Foundation Models

4

The Design Space of Tri-Modal Masked Diffusion Models

5

Mix, Don't Tune: Bilingual Pre-Training Outperforms Hyperparameter Search in Data-Constrained Settings

6

Large Language Models for Machine Translation Quality Annotation: Humans and Models Are Both Challenged

7

Minions: A Dynamic and Distributed Virtual GPU Cloud Built on Idle Apple Silicon Devices

8

Self-Compacting Language Model Agents

9

Normalizing Trajectory Models

10

RISED: Rubrics for Agentic Multi-Environment Selection and Self-Distillation

11

Negotiating Ontological Boundaries in User-Authored Personal Sensing Systems

12

Language Discrimination Improves Linguistic Learning in Multilingual Speech Models

13

How Much of a Harness Does a Strong Agent Need for Autonomous ML Engineering?

14

RLTL;DR: Self-Improvement by Internalizing Self-Generated Feedback

15

On the Effectiveness-Fluency Trade-Off in LLM Conditioning: A Systematic Study

16

The Communication Bottleneck: A Round-Trip Study of Tree-Structured Expression Serialization in Language Models

17

Faster Rates for Federated Variational Inequalities

18

A Practical Recipe for Semi-Supervised Federated ASR: Online Pseudo-Labels with Server Update Stabilization

19

Compressing Streaming Neural Audio Encoders via Latent-Space Distillation

20

How to Guide Your Language Flow

21

Dynamically Scaled Activation Steering

22

REVERSAL-BENCH: A Reversibility Axis and Reset Oracle for Measuring the Reset-Free RL Cliff

23

DACA-GRPO: Denoising-Aware Credit Assignment for Reinforcement Learning in Diffusion Language Models

24

Glyph: A Multi-Strategy Agentic System for Column Description and Sensitivity-Ontology Tagging of Enterprise Data Catalogs

25

How Value Induction Reshapes LLM Behaviour

26

Shared Selective Persistent Memory for Agentic LLM Systems

27

DiscoSign: Discourse-Aware Text to Sign Language Gloss Translation

28

Putting Captions to the Test: Evaluating Video Caption Quality through Multiple-Choice Question Answering

29

SimpleDesign: A Joint Model for Protein Sequence and Structure Codesign

30

REFACTOR-VLA: Unsupervised Library Learning of Typed Motor Programs

31

Agent Seer: Synthesizing Scenarios from Specification Understanding

32

LLMs Are Not (Consistently) Bayesian: Quantifying Internal (In)consistencies of LLMs’ Probabilistic Beliefs

33

From Preferences to Principles: Rubric-Based Alignment for Grounded Knowledge Answers

34

IDEA Prune: An Integrated Enlarge-and-Prune Pipeline in Generative Language Model Pretraining

35

Luce: Relightable Gaussians for 3D Asset Generation

36

PROOF-Gen: From Optimized Data to Better Distillation