Getting some extra storage space on your Mac could be as easy as deleting Apple's AI features with a new open-source command line tool called…
OpenAI's latest ad format will put images of sponsored products and services on your screen. The ads, which OpenAI will begin testing in the US…
Breast cancer is the most commonly diagnosed cancer among American women — yet the gaps in care are wide. A majority of women over age 40 skip the recommended annual screening. Radiologists are reading more mammograms with fewer colleagues. And when a diagnosis arrives, the tests that inform treatment can take weeks to return results. […]
Safeworld is building digital humans to make sure robots don't hurt the real ones.
New text-classification model. Tags: gemma4, image-text-to-text, system-one, system-two, adaptive-thinking, typed-decisions, decision-model, calibrated-probabilities
OpenAI introduces a new visual ad format in ChatGPT and expands measurement tools, attribution partnerships, and brand suitability for advertisers.
Norbert Wiener, godfather of cybernetics, once said, "The thought of every age is reflected in its technique." For the past century, our thought has been…
New image-text-to-text model. Tags: qwen3_5, image-text-to-text, jev, system-one, system-two, typed-decisions, calibrated-probabilities, multimodal
New automatic-speech-recognition model. Tags: cactus-needle, whistle, speech-recognition, speech-to-text, on-device, edge, quantization, webassembly
v0.31.0 Highlights This release features 717 commits from 307 contributors (96 new)! DeepSeek-V4.1-Flash performance: FlashMLA mega attention with the V4.1 NVFP4 compressed KV cache is now the SM100 default (56935); DeepGEMM sparse MQA logits for the indexer (56254) and Mega-Gate fusing the gate GEMM with expert selection (56266); decoder boundaries fuse the TP all-reduce, mHC input preparation (57643) and the…
A long-standing goal of AI is a model that can continually learn and improve itself. On post-trained models, supervised finetuning (SFT) on new data often causes poor generalization and catastrophic forgetting. As such, the conventional wisdom is that on-policy training is a prerequisite for continual learning. In practice, however, data containing new knowledge or capabilities are often off-policy. While methods such as on-policy self-distillation (OPSD) try to bridge this gap by converting off-policy data into on-policy signal, they have been shown to cause reasoning collapse. In this paper, we show that off-policy merging beats OPSD for continual learning. We first show that SFT learns a useful signal from new data, but naively applying its update interferes with existing capabilities. We reduce this interference with a simple recipe we term grafting, which changes where the update is learned and how it is applied: (1) learning the update on an earlier donor checkpoint, ideally even before the end of pretraining, and applying the weight update to the post-trained model; (2) scaling the weight update, equivalent to a form of model merging; and (3) optionally, masking the most sensitive update directions when the new data distribution is far from the post-trained model. Across continual learning settings including (1) distilling from expert traces, (2) self-improvement with STaR and Pedagogical RL, and (3) injecting knowledge after pretraining cutoff, grafting Pareto-dominates both SFT and OPSD in new-task and old-task performance, while avoiding expensive on-policy sampling. Therefore, our work challenges on-policy training as a necessity for continual learning on RL-trained models.
