Research map / Architectures and efficiency / Quantization
QuIP: 2-Bit Quantization of Large Language Models With Guarantees
Jerry Chee, Yaohui Cai, Volodymyr Kuleshov, Christopher De Sa
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add incoherence processing (LDLQ to QuIP) at 2 bits (ablation) | LDLQ/OPTQ · WikiText2 OPT-30b | perplexity: 71.70 -> 11.48 | Improved |
| add incoherence processing (LDLQ-RG to QuIP-RG) at 2 bits (ablation) | LDLQ-RG · WikiText2 OPT-30b | perplexity: 49.40 -> 11.68 | Improved |
| add incoherence processing to greedy updates at 2 bits (ablation) | Greedy · WikiText2 OPT-30b | perplexity: 4,817 -> 11.59 | Improved |
| add incoherence processing to nearest rounding at 2 bits (ablation) | Near · WikiText2 OPT-30b | perplexity: 41,548 -> 12.04 | Improved |
| add incoherence processing at 2 bits (ablation) | OPTQ · WikiText2 OPT-125m | perplexity: 4,563 -> 347.4 | Improved |
| add incoherence processing at 3 bits (ablation) | OPTQ · WikiText2 OPT-6.7b | perplexity: 14.87 -> 11.51 | Improved |
| add incoherence processing at 3 bits (ablation) | OPTQ · WikiText2 OPT-2.7b | perplexity: 17.09 -> 17.44 | Hurt |
| add incoherence processing at 4 bits (ablation) | OPTQ · WikiText2 OPT-125m | perplexity: 31.44 -> 33.35 | Hurt |
| add incoherence processing at 4 bits (ablation) | OPTQ · WikiText2 OPT-2.7b | perplexity: 12.93 -> 12.39 | Improved |
| replace nearest rounding with unbiased stochastic rounding in LDLQ/OPTQ (ablation) | LDLQ/OPTQ · OPT 125m, baseline processing 2-bit | perplexity difference: 241.3 | Hurt |
| use greedy updates alone instead of LDLQ in proxy loss at 2 bits (ablation) | Greedy · proxy loss | proxy loss: 2,554.89 -> 2,587.17 | No clear effect |
| use nearest rounding instead of LDLQ in proxy loss at 2 bits (ablation) | Near · proxy loss | proxy loss: 2,554.89 -> 5,971.69 | Hurt |
Later papers that compared with this method
19 later accepted papers report QuIP as a baseline in their results tables; 19 of them report a better number on most of the results they share with it.
- Structured Transforms for Low-Overhead Quantization of Language Models (NeurIPS 2026 · better on 7 of 9 shared results)
- UniSVQ: 2-bit Unified Scalar-Vector Quantization (ICML 2026 · better on 5 of 7 shared results)
- Proteus: Lookup-Free Trellis-Coded Quantization by Lattice-Breaking Compute Codes for 2-Bit LLMs (ICML 2026 · better on 6 of 6 shared results)
- GSQ: Highly-Accurate Low-Precision Scalar Quantization for LLMs via Gumbel-Softmax Sampling (NeurIPS 2026 · better on 2 of 2 shared results)
- PocketLLM: Ultimate Compression of Large Language Models via Meta Networks (AAAI 2026 · better on 2 of 2 shared results)
- OSAQ: Outlier Self-Absorption for Accurate Low-bit LLM Quantization (ICML 2026 · better on 1 of 1 shared result)
- RaBiT: Residual Aware Binarization Training for Accurate and Efficient LLMs (ICML 2026 · better on 1 of 1 shared result)
- Highly Efficient and Effective LLMs with Multi-Boolean Architectures (ICLR 2026 · better on 1 of 1 shared result)
11 more not listed.
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- QuIP$\#$: Even Better LLM Quantization with Hadamard Incoherence and Lattice Codebooks (ICML 2024 · similarity 0.92)
- Norm Tweaking: High-Performance Low-Bit Quantization of Large Language Models (AAAI 2024 · similarity 0.92)
- What Makes Quantization for Large Language Model Hard? An Empirical Study from the Lens of Perturbation (AAAI 2024 · similarity 0.91)
- QLLM: Accurate and Efficient Low-Bitwidth Quantization for Large Language Models (ICLR 2024 · similarity 0.91)
- Structured Transforms for Low-Overhead Quantization of Language Models (NeurIPS 2026 · similarity 0.91)
- BiLLM: Pushing the Limit of Post-Training Quantization for LLMs (ICML 2024 · similarity 0.91)
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.
