Skip to content
AI News HubLIVE
Public articles 137Collected articles 149Trust 90Refresh 30 min
Health HealthySource type ResearchFull-text rights Official full textLast ingested 2026-11-20ID apple-ml-researchStatus Enabled

Official research source; confirm reuse terms before enabling full body display.

Latest public articles

Split-and-Scale Latent 3D Representations

Apple researchers introduce split-and-scale, a method that treats 3D scenes as continuous functions of 3D coordinates so a single pretrained tokenizer can be applied to arbitrarily small sub-regions at inference time, expanding latent capacity and detail without raising training costs.

Apple Machine Learning ResearchIn-site articleSplit-and-Scale Latent 3D Representations

Theoretical Limits of Language Model Alignment

This paper characterizes the information-theoretic limits of KL-regularized alignment, deriving a closed-form expression for optimal reward improvement governed by Jeffreys divergence rather than √KL, and shows best-of-N approaches the limit while PPO and GRPO are suboptimal.

Apple Machine Learning ResearchIn-site articleTheoretical Limits of Language Model Alignment

Complexity of Differentially Private Selection via Federated APIs

Researchers show that private selection built on the standard noisy aggregation primitive inherently needs Ω(√d/ε) samples, and propose a new SVT-based primitive that yields an O(log(dn)/ε) additive-error selection algorithm, establishing an exponential separation from shuffle and aggregation models.

Apple Machine Learning ResearchIn-site articleComplexity of Differentially Private Selection via Federated APIs

Toward a Theory of Evaluability

This paper formalizes a notion of evaluability without the held-out assumption, investigating whether black-box models can be accurately evaluated when test samples may have been seen during training. It fully characterizes when accurate evaluation is possible for output-only access and shows that additional query access can enhance evaluator power, linking to self-correctors in program testing.

Apple Machine Learning ResearchIn-site articleToward a Theory of Evaluability

Weighted State Souping for Efficient Long-Context Retrieval

This paper introduces weighted state souping, which learns query-dependent weights to combine cached state encodings from state space models, substantially improving long-context retrieval, multi-hop QA, variable tracking, and tool use. A sparse variant achieves near-constant encoding cost with respect to context length. The work was accepted at the NeurIPS 2026 PALM workshop.

Apple Machine Learning ResearchIn-site articleWeighted State Souping for Efficient Long-Context Retrieval

On the Importance of Gating: Memorization vs. In-Context Learning in State Space Models

This work shows that the gating mechanism in State Space Models causes them to first learn an in-weights memorization solution, delaying or preventing convergence to correct in-context learning, even without architectural limitations. However, gating often benefits generalization to long sequence lengths.

Apple Machine Learning ResearchIn-site articleOn the Importance of Gating: Memorization vs. In-Context Learning in State Space Models

Evaluating the Evaluators: Investigating LLM Judges for Personalized Writing Style Assessment

Researchers from Apple and collaborating institutions evaluate LLM judges for personalized writing style tasks. They find judge quality correlates only weakly with general LLM ability (Pearson ρ ≤ 0.2), varies by writing task and evaluation criteria, and that paired comparisons outperform independent scoring. Notably, the strongest LLMs do not most reliably reproduce human evaluations.

Apple Machine Learning ResearchIn-site articleEvaluating the Evaluators: Investigating LLM Judges for Personalized Writing Style Assessment

Faster Retrieval for Better, More Efficient RAG

Apple researchers present a method at ICML 2026 that trains small neural networks to learn support functions for maximum inner product search, speeding up retrieval in RAG and other applications while maintaining high recall.

Apple Machine Learning ResearchIn-site articleFaster Retrieval for Better, More Efficient RAG

Attention to Mamba: A Recipe for Cross-Architecture Distillation

Apple researchers propose a two-stage distillation method that first distills a Transformer into a linearized attention model and then into a Mamba architecture, achieving a perplexity of 14.01 close to the teacher's 13.86 on Pythia-1B and largely preserving downstream performance, offering a new pathway for adopting SSMs.

Apple Machine Learning ResearchIn-site articleAttention to Mamba: A Recipe for Cross-Architecture Distillation

Online Dense Reinforcement Learning with Planning-Grounding Agent for UI Navigation on Live iOS Apps

Apple researchers propose an online reinforcement learning framework that trains UI navigation agents directly on live iOS applications. The framework decouples the agent into a planning agent and a UI grounding model, uses an LLM-as-a-Judge to provide dense step-level and terminal rewards, and optimizes with PPO and step-level advantage estimation. They introduce GiGPPO for critic stability and an infrastructure orchestrating hundreds of concurrent virtual iOS devices. Their 8B RL agents achieve nearly a sevenfold improvement over SFT baselines in out-of-distribution generalization to Android and outperform Gemini-3.5-Flash while tying Gemini-3.1-Pro on real iOS apps.

Apple Machine Learning ResearchIn-site articleOnline Dense Reinforcement Learning with Planning-Grounding Agent for UI Navigation on Live iOS Apps

GHARP: Real-Time Gaussian Head Animation from Large-Scale Reconstruction Prior

GHARP animates 3D human heads in real time from a few input images and a driving expression signal. It decouples the problem into an offline identity stage and a lightweight runtime animation stage, and uses a body alignment network to resolve ambiguity from body pose and clothing, reducing blur and flicker. It achieves state-of-the-art quality on Ava-256 while running up to 13× faster with 8× fewer Gaussians.

Apple Machine Learning ResearchIn-site articleGHARP: Real-Time Gaussian Head Animation from Large-Scale Reconstruction Prior

Pixels to Tokens: Token Space Efficient Active Learning for Low-Budget Semantic Segmentation

TEAL is the first pixel-level active learning framework built on ViT representations, performing diversity-based selection directly on the native token lattice and refining with margin-based uncertainty on a finer decoder grid. It improves up to +24.36 mIoU over the strongest baseline with one labeled pixel per image and reaches 97% of fully supervised performance with just 10 pixels.

Apple Machine Learning ResearchIn-site articlePixels to Tokens: Token Space Efficient Active Learning for Low-Budget Semantic Segmentation

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

RISED uses LLM-generated rubrics to guide data selection and self-distillation when training a single agent across multiple interactive environments. It improves mean pass rates and per-environment performance over baselines.

Apple Machine Learning ResearchIn-site articleRISED: Rubrics for Agentic Multi-Environment Selection and Self-Distillation

All sources