Unverified50% confidenceFactExact time
Speculative decoding has spawned multiple variants including Medusa (multi-head prediction), EAGLE (generating drafts using hidden states from early layers of the main model), and DFlash (hybrid layout optimization).
1
Sources
50%
Confidence
Long-term
Relevance
9/6/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedMedusa adds multiple prediction heads on top of a model to predict tokens at different future positions and confirms them through tree-based verification, without requiring an additional draft model, and can improve inference speed by 2–3×.72% similarVerifiedSpeculative Decoding uses a small, fast draft model to predict multiple tokens in advance, which are then verified in parallel all at once by the large model.68% similarUnverifiedAnthropic's sparse autoencoder research has been able to isolate millions of interpretable features from model activations65% similarUnverifiedDPO通过数学推导绕过奖励模型，直接在偏好数据上用分类损失函数优化语言模型，被Llama 3、Mistral等开源模型广泛采用64% similarUnverifiedStructured Outputs通过约束解码（Constrained Decoding）技术实现，在模型生成token时动态限制可选token的范围64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/864585API
curl https://kongchang.com/api/v1/knowledge/claims/864585MCP
get_claim(id=864585)