This paper explores how privileged information enhances on-policy self-distillation (OPSD) in language models by comparing different reasoning views on a new dataset of mathematical problems. The study shows that while reference-free distillation contributes significantly to model improvements, the additional benefit from privileged references is limited and depends on the student's training state.
The chart will appear after repeat observations. The current metric comes from the source.
2 source votesReal observations only. History before source connection is not reconstructed.
This research helps understand how incorporating privileged information affects self-distillation in language models, which can guide the design of more effective training strategies.
LimiX-2 is a new model in the LimiX family that uses the Contextual Mechanism Networks (CMNs) paradigm, pretrained with Context-Conditional Masked Modeling (CCMM). It focuses on learning joint representations of data generation processes rather than target-centric prediction, showing superior performance on tabular data benchmarks and promoting causal awareness.
3 days agoDaily PapersStepAudio 3 Realtime is an audio-language foundation model designed for real-time spoken interaction, featuring a continuous listen-converse-think-act loop. It includes Deep Perception for interpreting user intent, Seamless Duplex for handling audio stream synchronization, and Think-While-Speaking to balance reasoning and latency. The model achieves high performance on benchmarks like MMSU and Artificial Analysis Full-Duplex, along with a task-success rate on τ-Voice.
6 days agoDaily PapersPhysBrain 1.5 is a unified model for understanding physical environments, generating actions, and predicting future states. It uses a vision-language foundation, encodes language responses and motion as sequences, and is pre-trained on human interaction videos. The model achieves strong performance on embodied understanding benchmarks.
4 days agoDaily PapersThe paper introduces XConf, a method for estimating confidence in language models by leveraging the model's accumulated experience. XConf uses past episodes, including tasks, reflections, confidence levels, outcomes, and lessons learned, to inform current confidence estimates through a recall and reflect process. It outperforms existing methods in discrimination and calibration with lower computational cost.
3 days agoDaily PapersLynnReal-Omni is a native multimodal video generation framework that combines text-to-video, image-conditioned generation, and structural control into a single model. It includes a 32B shared multimodal diffusion transformer and a 27B Flash version for real-time rendering, with a data pipeline and evaluation method designed for video generation tasks.
4 days agoDaily PapersThe paper introduces ScienceIDE, a framework that transforms scientific code repositories into programmable environments for training scientific agents. These environments enable task generation, execution, and verification, supporting supervised fine-tuning and reinforcement learning. The approach demonstrates improvements in scientific code repair and general-purpose benchmarks.
2 days ago