paperarXivTrust 82 · PrimaryPublished 3mo agoLive · 2mo ago
Quantization at 1.58 bits
Ternary-weight models that retain most of full-precision quality.
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- LinkedLinked via unknownQuantization →
- LinkedLinked via unknownquant-kit →
- LinkedLinked via unknownQuantBench →
- PossiblePossibly related (embedding) · 51%Picovoice/picollm →
- PossiblePossibly related (embedding) · 47%ExTernD: Expanded-Rank Ternary Decomposition Ternary LLM PTQ with Accuracy Approaching Any Quantization Level [P] →