FeedSources
AM

Apple Machine Learning Research Publications

47 articles total

Go to source

Explore advancements in state of the art machine learning research in speech and natural language, privacy, computer vision, health, and more.

Go to source
  •  GH-ESD: Grounded Hypothesis-Driven Error Slice Discovery for Instance-Level Vision Tasks
  •  LEAD: Breaking the No-Recovery Bottleneck in Long-Horizon Reasoning
  •  Accelerating Text-to-Video Generation with Calibrated Sparse Attention
  •  Environment-free Synthetic Data Generation for API-Calling Agents
  •  Length Value Model: Scalable Value Pretraining for Token-Level Length Modeling
  •  LVSum: A Benchmark for Timestamp-Aware Long Video Summarization
  •  RayRoPE: Projective Ray Positional Encoding for Multi-View Attention
  •  Show Me Examples: Inferring Visual Concepts from Image Sets
  •  When Unlearning Is Free: Leveraging Low Influence Points to Reduce Computational Costs
  •  Embarrassingly Simple Self-Distillation Improves Code Generation
  •  Personalizing Incremental Video Search with Hybrid Text and ID Embeddings
  •  Doubly Sub-linear Interactive Proofs of Proximity
  •  Interactive Proofs for General Distribution Properties
  •  Location-Invariant Properties of Functions Versus Properties of Distributions: United in Testing but Separated in Verification
  •  CLaRa: Bridging Retrieval and Generation with Continuous Latent Reasoning
  •  One Layer Is Enough: Adapting Pretrained Visual Encoders for Image Generation
  •  Uncertainty Quantification for LLM Function-Calling
  •  Multilingual Semantic Retrieval for Apple Music Search
  •  Proactive Agent Research Environment: Simulating Active Users to Evaluate Proactive Assistants
  •  Behavioral Privacy Leakage in Agentic Negotiation: Formalizing and Mitigating Inference Attacks via Randomized Policies
  •  Recursive Language Models Meet Uncertainty: The Surprising Effectiveness of Self-Reflective Program Search for Long Context
  •  Unmasking On-Policy Distillation: Where It Helps, Where It Hurts, and Why
  •  Incentivizing Temporal-Awareness in Egocentric Video Understanding Models
  •  A Single Neuron Is Sufficient to Bypass Safety Alignment in Large Language Models
  •  DynaMiCS: Fine-Tuning LLMs with Performance Constraints Using Dynamic Mixtures
  •  FlowEval: Reference-Based Evaluation of Generated User Interfaces
  •  LensVLM: Selective Context Expansion for Compressed Visual Representation of Text
  •  MT-EditFlow: Reinforcement Learning for Multi-Turn Image Editing with Flow Matching
  •  Taming Text-to-Sounding Video Generation via Advanced Modality Condition and Interaction
  •  Fortress: A Case Study in Stabilizing Search Recommendations via Temporal Data Augmentation and Feature Pruning
  •  Path-Constrained Mixture-of-Experts
  •  Revisiting ASR Error Correction with Specialized Models
  •  Scaling Properties of Continuous Diffusion Spoken Language Models
  •  Segmental Attention Decoding with Long Form Acoustic Encodings
  •  TopoPrimer: The Missing Topological Context in Forecasting Models
  •  (1D) Ordered Tokens Enable Efficient Test-Time Search
  •  Amortizing Maximum Inner Product Search with Learned Support Functions
  •  Annotations Mitigate Post-Training Mode Collapse
  •  Anti-Causal Domain Generalization: Leveraging Unlabeled Data
  •  Balanced LoRA: Removing Parameter Invariance to Accelerate Convergence