Research map / Architectures and efficiency / Attention and transformers
Transformers in Pseudo-Random Number Generation: A Dual Perspective on Theory and Practice
Ran Li, Lingshu Zeng
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 |
|---|---|---|---|
| train GPT-2 on MT19937 output at 8, 12 and 16 bits | GPT-2 · MT19937 simulation | training accuracy: approximately 1.0 at all bit-widths | Improved |
| increase bit-width from 8 to 16 (ablation) | GPT-2 · MT19937 simulation | convergence steps: 500 steps -> 1,000 -> 2,000 steps | Hurt |
| generate from Transformer checkpoints at accuracy 0.63-0.89 | GPT-2 PRNG · NIST STS | tests passed: 11 of 15 passed | Mixed |
Closest papers by meaning
- (How) Can Transformers Predict Pseudo-Random Numbers? (ICML 2025 · similarity 0.87)
- Algorithmic Capabilities of Random Transformers (NeurIPS 2024 · similarity 0.86)
- The Expressive Power of Transformers with Chain of Thought (ICLR 2024 · similarity 0.83)
- Transformers need glasses! Information over-squashing in language tasks (NeurIPS 2024 · similarity 0.83)
- Understanding Transformers via N-Gram Statistics (NeurIPS 2024 · similarity 0.82)
- Gumiho: A Hybrid Architecture to Prioritize Early Tokens in Speculative Decoding (ICML 2025 · similarity 0.82)
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.
