Skip to main content
Context Awesome

awesome-speculative-decoding

github.com/geralt-targaryen/awesome-speculative-decoding ↗

Reading notes on Speculative Decoding papers

36
GitHub Stars
141
Curated Resources
8
Categories
21 hours ago
Last Refreshed
Bibliography by VenuesHistory & OriginDraft ModelsRetrieval-based Speculative DecodingDraft Tree ConstructionVerification StrategiesDraft Length ControlSpeculative Decoding + Other Technologies

Use this list with your AI agent

Add the Context Awesome MCP server to Claude, Cursor, or any MCP client, then ask:

"Show me iclr 2026 resources from awesome-speculative-decoding"

Installation instructions →

What's inside

Bibliography by Venues

Retrieval-based Speculative Decoding

Draft Length Control

Draft Models

Verification Strategies

  • paper

  • paper

    Optimal Transport with Membership cost - and an approximation that's linear in vocabulary size and logarithmic in candidate set size to tackle draft selection when there are multiple drafts

Draft Tree Construction

Speculative Decoding + Other Technologies

History & Origin

Showing a sample of 141 resources. View the full list on GitHub →