Starts Nov 23Live Cohort 3 of Vibe Coding & AI Agents: 3 weeks, 9 live hands-on sessionsCohort 3 (Nov 23) →
Publisher · verified by editors

Apple ML Research

AI news source. Articles are auto-selected and adapted by Hamidun News editors.

26 articles in Hamidun·Latest: October 13· Active·machinelearning.apple.com ↗

Latest publications

Apple proposes SelfCompact: the agent decides for itself when to compress context
ResearchApple ML Research

Apple proposes SelfCompact: the agent decides for itself when to compress context

Apple researchers removed the fixed token threshold from context compression: in SelfCompact, the model chooses when and how to compact its context through a dedicated tool.

Oct 13, 2026·2 min
Apple Research Unveils Minions: A Virtual GPU Cloud Built on Idle Macs
HardwareApple ML Research

Apple Research Unveils Minions: A Virtual GPU Cloud Built on Idle Macs

Apple researchers describe Minions, a distributed cloud that pools idle Apple Silicon GPUs in macOS devices to run ML inference at scale.

Oct 13, 2026·2 min
Apple tested LLMs as translation judges: MQM and ESA are hard for both humans and models
ResearchApple ML Research

Apple tested LLMs as translation judges: MQM and ESA are hard for both humans and models

Apple ML Research compared LLM and human quality annotations for machine translation under the MQM and ESA schemes. The title gives the verdict: both sides struggle.

Oct 13, 2026·2 min
Apple ML Research Studies Alignment in Multimodal LLMs and Their Tendency to Hallucinate
CreativeApple ML Research

Apple ML Research Studies Alignment in Multimodal LLMs and Their Tendency to Hallucinate

Apple ML Research has published a broad study of preference alignment in multimodal language models, and of why hallucinations there are more complex than in text-only LLMs.

Aug 4, 2026·2 min
Apple ML Research proposes compact models to fix speech recognition errors
LLMApple ML Research

Apple ML Research proposes compact models to fix speech recognition errors

Apple ML Research proposes compact seq2seq models instead of LLMs to correct ASR errors, training them on real and synthetic errors.

Jul 19, 2026·2 min
Apple ML Research presents Text-to-Sounding-Video, a method for generating video with sound from text
CreativeApple ML Research

Apple ML Research presents Text-to-Sounding-Video, a method for generating video with sound from text

Apple ML Research outlined Text-to-Sounding-Video, which generates video with sound from text, and named two key unsolved problems in the field.

Jul 19, 2026·2 min
Apple ML Research unveils TopoPrimer, a topological context layer for forecasting models
ResearchApple ML Research

Apple ML Research unveils TopoPrimer, a topological context layer for forecasting models

Apple ML Research has introduced TopoPrimer, a framework that feeds the global topological structure of a dataset into time-series forecasting models.

Jul 18, 2026·1 min
Apple works out when on-policy distillation actually helps model training
LLMApple ML Research

Apple works out when on-policy distillation actually helps model training

Apple ML researchers identify the conditions under which on-policy distillation works for training reasoning models, and when it can be ineffective or even counterproductive.

Jul 13, 2026·3 min
Self-Reflective Program Search: Apple Improves How LLMs Handle Long Contexts
LLMApple ML Research

Self-Reflective Program Search: Apple Improves How LLMs Handle Long Contexts

Apple ML Research has published a study of Self-Reflective Program Search, a method that improves language models' handling of long contexts by recursively breaking them into subqueries.

Jul 13, 2026·2 min
Apple develops TGPO to teach video models to understand time
CreativeApple ML Research

Apple develops TGPO to teach video models to understand time

Apple ML Research unveils TGPO, a method that teaches video models the order and evolution of events in egocentric video by reinforcing temporal reasoning.

Jul 13, 2026·2 min
AI Vulnerability: A Single Neuron Can Bypass Safety Guardrails in Large Language Models, Apple Finds
LLMApple ML Research

AI Vulnerability: A Single Neuron Can Bypass Safety Guardrails in Large Language Models, Apple Finds

Apple researchers show that LLM safety protections can be fully bypassed by manipulating a single neuron, with no fine-tuning or prompt changes, across seven models from two families.

Jul 9, 2026·2 min
Apple's FlowEval Uses AI to Grade AI-Generated Interfaces
CreativeApple ML Research

Apple's FlowEval Uses AI to Grade AI-Generated Interfaces

Apple ML Research has built FlowEval, a framework that automatically scores AI-generated interfaces by comparing their navigation trajectories with those of real websites.

Jul 9, 2026·2 min
Apple ML Research unveils Weblica, a scalable environment for training visual web agents
LLMApple ML Research

Apple ML Research unveils Weblica, a scalable environment for training visual web agents

Apple has released Weblica, a framework that replays interactive web page states and uses LLMs to generate training environments, aiming to train browser-based AI agents reliably.

Jul 9, 2026·2 min
Apple develops MT-EditFlow for multi-step AI image editing
CreativeApple ML Research

Apple develops MT-EditFlow for multi-step AI image editing

Apple ML Research has unveiled MT-EditFlow, a reinforcement learning system that tackles the problem of errors piling up during iterative image editing.

Jul 9, 2026·2 min
LensVLM: A Neural Network for Reading Text in Compressed Images
CreativeApple ML Research

LensVLM: A Neural Network for Reading Text in Compressed Images

Apple ML Research has proposed LensVLM, a framework that lets vision-language models read small text in compressed images without losing accuracy.

Jul 8, 2026·3 min
How to Fine-Tune an AI Model on New Tasks Without Losing Its Existing Abilities
LLMApple ML Research

How to Fine-Tune an AI Model on New Tasks Without Losing Its Existing Abilities

Apple's DynaMiCS treats data selection during fine-tuning as a constrained problem, preserving a model's instruction-following, knowledge and safety while it adapts to new domains.

Jul 8, 2026·2 min
How Mixture-of-Experts Works: The Architecture Behind Modern Neural Networks
ResearchApple ML Research

How Mixture-of-Experts Works: The Architecture Behind Modern Neural Networks

Apple researchers found that tokens in MoE architectures use only a tiny fraction of the theoretically possible routes, and they propose an architecture to fix it.

Jul 8, 2026·3 min
Speech neural networks scale like text models, Apple study finds
CreativeApple ML Research

Speech neural networks scale like text models, Apple study finds

Apple ML Research found that speech language models built on continuous diffusion scale as predictably as autoregressive ones, and introduced a new metric, pJSD.

Jul 8, 2026·2 min
Apple ML Research tackles ASR accuracy loss on long audio recordings
ResearchApple ML Research

Apple ML Research tackles ASR accuracy loss on long audio recordings

Apple ML Research has published Segmental Attention Decoding, four architectural changes aimed at the main problem neural networks face with long audio recordings.

Jul 8, 2026·2 min
Apple publishes Fortress, a framework that stabilizes recommender systems by pruning features
ResearchApple ML Research

Apple publishes Fortress, a framework that stabilizes recommender systems by pruning features

Apple ML Research describes Fortress, a system that removes temporal instability from recommendation models by finding and deleting features that produce inconsistent predictions.

Jul 8, 2026·2 min
Apple ML Research examines why AI safety data annotators disagree
MedicineApple ML Research

Apple ML Research examines why AI safety data annotators disagree

Apple ML Research has developed a method for analyzing disagreements among safety data annotators, the people who train AI models to tell acceptable content from dangerous content.

Jul 8, 2026·2 min
Apple ML Research proposes a way to generalize ML models to new domains without labeling
ResearchApple ML Research

Apple ML Research proposes a way to generalize ML models to new domains without labeling

Apple ML Research has developed a domain generalization approach that relies on unlabeled data from the new domain instead of costly annotation.

Jul 4, 2026·2 min
A More Efficient Neural Network: How Apple Optimized Inference
LLMApple ML Research

A More Efficient Neural Network: How Apple Optimized Inference

Apple ML Research unveils Conformal Thinking, a framework that automatically manages the token budget of reasoning models, guaranteeing a set error rate with minimal compute.

Jul 4, 2026·2 min
Apple ML Research: How Diffusion Language Models Can Learn to Pick Tokens Without Hand-Tuned Heuristics
CreativeApple ML Research

Apple ML Research: How Diffusion Language Models Can Learn to Pick Tokens Without Hand-Tuned Heuristics

Apple ML Research proposes replacing hand-built token-selection heuristics in diffusion language models with learned policies, aiming to remove instability and manual parameter tuning.

Jul 4, 2026·3 min
Apple ML Research proposes MemoryLLM, an interpretable "memory" for transformers
LLMApple ML Research

Apple ML Research proposes MemoryLLM, an interpretable "memory" for transformers

Apple researchers describe feed-forward blocks in LLMs as neural lookup memory and propose a way to analyze them in isolation, without the self-attention mechanism.

Jul 4, 2026·2 min
Apple ML Research: Multi-Agent LLM Teams Hold Back Their Expert Agents
LLMApple ML Research

Apple ML Research: Multi-Agent LLM Teams Hold Back Their Expert Agents

Apple ML Research finds that self-organizing teams of language models produce no synergy. They hold expert agents back instead of boosting them.

Jul 4, 2026·3 min