zetabyte

​We present StreamBridge, a simple yet effective framework that seamlessly transforms offline Video-LLMs into streaming-capable models. It addresses two fundamental challenges in adapting existing models into online scenarios: (1) limited capability for multi-turn real-time understanding, and (2) lack of proactive response mechanisms. Specifically, StreamBridge incorporates… Read More »StreamBridge: Turning Your Offline Video Large Language Model into a Proactive Streaming Assistant Apple Machine Learning Research

​Language models must be adapted to understand and follow user instructions. Reinforcement learning is widely used to facilitate this — typically using fixed criteria such as “helpfulness” and “harmfulness”. In our work, we instead propose using flexible, instruction-specific criteria as a means of broadening the… Read More »Checklists Are Better Than Reward Models For Aligning Language Models Apple Machine Learning Research

​Recent generations of frontier language models have introduced Large Reasoning Models (LRMs) that generate detailed thinking processes before providing answers. While these models demonstrate improved performance on reasoning benchmarks, their fundamental capabilities, scaling properties, and limitations remain insufficiently understood. Current evaluations primarily focus on established… Read More »The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity Apple Machine Learning Research

​While federated learning (FL) and differential privacy (DP) have been extensively studied, their application to automatic speech recognition (ASR) remains largely unexplored due to the challenges in training large transformer models. Specifically, large models further exacerbate issues in FL as they are particularly susceptible to… Read More »Enabling Differentially Private Federated Learning for Speech Recognition: Benchmarks, Adaptive Optimizers, and Gradient Clipping Apple Machine Learning Research

​[[{“value”:”We revisit the problem of secure aggregation of high-dimensional vectors in a two-server system such as Prio. These systems are typically used to aggregate vectors such as gradients in private federated learning, where the aggregate itself is protected via noise addition to ensure differential privacy.… Read More »PREAMBLE: Private and Efficient Aggregation via Block Sparse Vectors Apple Machine Learning Research

​Self-supervised learning (SSL) on graphs generates node and graph representations (i.e., embeddings) that can be used for downstream tasks such as node classification, node clustering, and link prediction. Graph SSL is particularly useful in scenarios with limited or no labeled data. Existing SSL methods predominantly… Read More »Generative and Contrastive Graph Representation Learning Apple Machine Learning Research

​[[{“value”:” Can safety keep up with real-time LLMs? Alibaba’s Qwen team thinks so, and it just shipped Qwen3Guard—a multilingual guardrail model family built to moderate prompts and streaming responses in-real-time. Qwen3Guard comes in two variants: Qwen3Guard-Gen (a generative classifier that reads full prompt/response context) and… Read More »Meet Qwen3Guard: The Qwen3-based Multilingual Safety Guardrail Models Built for Global, Real-Time AI Safety Asif Razzaq Artificial Intelligence Category – MarkTechPost