DailyPapers

@HuggingPapers

Tweeting interesting papers submitted at huggingface.co/papers. Submit your own at hf.co/papers/submit, and link models/datasets/demos to it!

Anywhere

DailyPapers
@HuggingPapers
1h
Tencent releases AuK, a unified speech generation and editing model It handles zero-shot TTS, instruction-driven content/acoustic/paralinguistic editing, plus enhancement and separation through one natural-language interface. The distilled AuK-Flash version runs 4.5x faster.
1
DailyPapers
@HuggingPapers
4h
NVIDIA just released its Nemotron math specialist on Hugging Face Trained to solve hard olympiad problems and verify proofs Part of an ensemble that achieved a gold-medal-level score at IMO 2026
2
DailyPapers
@HuggingPapers
6h
Gander: a realtime omni interaction agent Unifies omni perception, full-duplex spoken interaction, and agentic execution in one open-source model, now on Hugging Face.
2
DailyPapers
@HuggingPapers
10h
NeoHorse-1: A step toward recursive self-improvement Agentic post-training with a routing harness turns execution traces into training data, narrowing the gap between a 4B model and a 9B baseline by almost 6 points across 10 benchmarks.
3
DailyPapers
@HuggingPapers
13h
NVIDIA just released its olympiad-grade math reasoning model on Hugging Face This model excels at solving and verifying mathematical proofs, and was part of the system that hit gold-medal level at IMO 2026
2