Unverified50% confidenceSolutionExact time
上下文偏置/热词偏置在解码阶段通过WFST或注意力机制动态提升特定词汇的输出概率，实现方式包括浅层融合和深层偏置
1
Sources
50%
Confidence
Long-term
Relevance
8/21/2026
First Seen
Sources
Related Claims
UnverifiedIn the denoising process of diffusion models, text prompts guide the generation direction through the Cross-Attention mechanism, with each keyword competing for weight in the attention layers.71% similarUnverifiedOffline file transcription can use bidirectional attention mechanisms to integrate information from both preceding and following context, and can perform multiple rounds of decoding and reranking to improve accuracy.69% similarUnverifiedFSBoard primarily covers fingerspelling rather than complete sign language vocabulary and grammatical expressions, leaving room for improvement in models trained on this dataset for understanding coherent sign language sentences.66% similarUnverifiedMidjourney uses CLIP's text encoder to convert prompts into conditioning vectors that guide the denoising process of the diffusion model.65% similarUnverifiedMOSS-Transcribe-Diarize has a built-in Hotword Boosting feature that improves recognition accuracy for proper nouns and code-switching (Chinese-English mixed) scenarios.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/782686API
curl https://kongchang.com/api/v1/knowledge/claims/782686MCP
get_claim(id=782686)