AAI News Hub

#distillation

10 briefs tagged #distillation.

Research

Study tests strong-to-weak transfer at inference time

A stronger model built task harnesses that lifted weaker-model scores on Theory-of-Mind benchmarks.

arXiv cs.AI+1 outlet·4d ago
Research

Paper proposes causal distillation for video generation

Context-Matched Distillation aligns teacher supervision with autoregressive video inference.

HF Daily Papers·4d ago
Research

Researchers shrink visual document retrieval models

DistilVDR uses dual-student distillation to build a 524M end-to-end visual document retriever.

arXiv cs.CL+1 outlet·5d ago
Research

Paper tests strong-to-weak AI scaffolding at inference

A stronger model built harnesses that lifted weaker models on Theory-of-Mind benchmarks without training.

HF Daily Papers+1 outlet·5d ago
Models

ByteDance trains large model to rival Anthropic

The Chinese tech giant is pre-training a model that could reach 10 trillion parameters.

Ars Technica AI+1 outlet·Aug 7
Research

OPD2 improves multilingual math reasoning in Qwen3 tests

The distillation method beat OPD in English, Korean and Japanese math reasoning experiments.

arXiv cs.LG+1 outlet·Aug 7
Research

Researchers refine on-policy distillation for smaller models

New arXiv papers propose capacity- and outcome-aware ways to guide student models.

arXiv cs.LG+1 outlet·Aug 7
Research

Researchers propose SKILL-KD for LLM agent skill distillation

The framework turns teacher-student trajectory gaps into tested textual skill patches.

arXiv cs.AI+1 outlet·Aug 5
Research

New papers probe privileged information in OPSD

Researchers test how rubrics, anchors and problem structure affect self-distilled model training.

arXiv cs.AI+1 outlet·Aug 4
Research

ReflectRL trains on failed expert reasoning traces

The paper proposes using flawed expert trajectories as reflection signals in on-policy training.

HF Daily Papers+1 outlet·Aug 4