Inkling Collection Inkling is a versatile, customizable model that reasons over text, images, audio, with variable and efficient thinking effort. • 2 items • Updated 9 days ago • 35
GigaChat 3.5 Collection GigaChat 3.5 is a large-scale Mixture-of-Experts (MoE) Hybrid model with 432B total parameters • 5 items • Updated 10 days ago • 18
Speculative Decoding Modules Collection A collection of speculative decoding modules created using Model Optimizer. • 18 items • Updated about 19 hours ago • 18
Nemotron-Post-Training-v3 Collection Collection of datasets used in the post-training phase of Nemotron Nano, Super, and Ultra v3. • 50 items • Updated 8 days ago • 181
When Search Agents Should Ask: DiscoBench for Clarification-Aware Deep Search Paper • 2606.27669 • Published 29 days ago • 15
JetSpec: Breaking the Scaling Ceiling of Speculative Decoding with Parallel Tree Drafting Paper • 2606.18394 • Published 30 days ago • 35
Deeper is Not Always Better: Mitigating the Alignment Tax via Confident Layer Decoding Paper • 2606.21906 • Published Jun 20 • 25
Qwen-RobotWorld Technical Report: Unifying Embodied World Modeling through Language-Conditioned Video Generation Paper • 2606.17030 • Published Jun 15 • 47
VIA-SD: Verification via Intra-Model Routing for Speculative Decoding Paper • 2606.12243 • Published Jun 10 • 36
Breaking Entropy Bounds: Accelerating RL Training via MTP with Rejection Sampling Paper • 2606.12370 • Published Jun 10 • 21
Multilingual LLM Evaluation Collection Multilingual Evaluation Benchmarks • 8 items • Updated Jul 31, 2025 • 36