AI Paper Atlas

Research map / Architectures and efficiency / Quantization

TurboBoA: Faster and Exact Attention-aware Quantization without Backpropagation

Junhan Kim, Yeo Jeong Park, Seungwoo Son, Chungman Lee, Ho-young Kim, Joonyoung Kim, Yongkweon Jeon

ICLR 2026 · Architectures and efficiency › Quantization · approach: Post-training quantization of LLMs · 3 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
quantize N=16 out-channels jointly instead of one by one (F1) (ablation)BoA · Quantization time, INT2 Llama3-8Btime (min): 94.75 -> 25.30Improved
quantize N=16 out-channels jointly (F1) (ablation)BoA · Wiki2 PPL, INT2 Llama3-8Bperplexity: 15.20 -> 15.41No clear effect
quantize N=64 jointly instead of N=16 (ablation)TurboBoA F1 · Quantization time Llama3-8Btime (min): 25.30 -> 21.56No clear effect
add error compensation for pre-quantized layers (F2) (ablation)TurboBoA · Wiki2 PPL INT2 Llama3.2-1Bperplexity: 41.85 -> 37.15Improved
add adaptive grid with CD refinement (F3) (ablation)TurboBoA · Wiki2 PPL INT2 Llama3.2-1Bperplexity: 41.85 -> 39.45Improved
add both F2 and F3 (ablation)TurboBoA · Wiki2 PPL INT2 Llama3.2-1Bperplexity: 41.85 -> 33.33Improved
add F3 (adaptive grid) to TurboBoA on Llama3.2-1B C4 (ablation)TurboBoA · C4 PPL INT2 Llama3.2-1Bperplexity: 108.1 -> 107.3Improved
add error compensation F2 (extra FP forward pass) (ablation)TurboBoA · quantization time, Llama3-8Btime (min): 25.30 -> 40.16Hurt
replace BoA with TurboBoA in weight-activation quantizationSpinQuant W2A4KV4 · Wiki2 PPLperplexity: 77.05 -> 63.07 (Llama3.2-1B)Improved
replace BoA with TurboBoA for 3-bit weight-onlyQuaRot INT3 · Wiki2 PPL Llama3-8Bperplexity: 7.145 -> 7.116No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy