Tag
A new hybrid decoding framework called In-Writing is proposed, which delays constraint application until after a trigger token, combining free-form reasoning with structured generation for improved accuracy in classification and reasoning tasks.
Researchers propose a training-free method called Suffix-Anchored Confidence Modulation to improve confidence-based decoding in diffusion language models by addressing issues with EOT tokens and premature decoding.
NVIDIA introduces PiD, a Pixel Diffusion Decoder that replaces traditional VAE/RAE decoders in latent diffusion models, enabling fast, high-resolution decoding with up to 6× speedup and improved visual fidelity.
A tweet recommending --ddtree-budget 36 for Nvidia RTX 4090, claiming 2.5x speedup during decoding for Qwen3.6_27B.
CoRD is a collaborative multi-teacher decoding framework that synthesizes reasoning trajectories through predictive perplexity scoring and beam search, enabling efficient distillation of large reasoning models with high-quality outputs and generalized performance.
A research paper proposing NWCAD (No-Worse Context-Aware Decoding), a decode-time adapter that prevents 'neutral regression' where LLMs overwrite already-correct answers when given non-informative context, using a two-stream architecture with gated fallback to no-context decoding.