Q1D-8B-Blk0-V1

Q1D 8B Blk0 V1

Research artifact from Q1 Descent — reconstructing per-block 1-bit "intelligence-density" recovery for open-weight Qwen3 models. Shelved 2026-06-19; published for the record. See the project debrief for context.

Method (Q1_0_g128, 1.125 bpw). Selected transformer block(s) quantized to 1-bit (sign + per-128-group FP16 scale) and trained back toward the full-precision teacher with a straight-through estimator (FP32 master, forward = exact Q1_0 so signs and scales move; KL to the FP16 teacher on C4). All other blocks stay F16. This is not a fully 1-bit model — it isolates how far a small number of 1-bit blocks can be healed.

This model. Block 0 only at 1-bit at the headline 8B scale; blocks 1-35 F16.

State / eval. GSM8K flex 0.70 / strict 0.51 (n=100; F16 teacher 0.82/0.77). Under-recovered at 8B vs 4B's near-parity. Single block, one seed.

Use. llama.cpp / LM Studio. Ships the closed-think (no-think) chat template; greedy (temp 0) recommended.

Caveats. Early research artifact, single seed, small-n evals (n=100, SE≈0.04 — don't over-read sub-0.08 gaps). Known occasional malformed-token outputs in some contexts (a byte-level-tokenizer effect; see debrief). Not affiliated with PrismML or the Qwen team.