Research map / Large language models / Instruction and fine-tuning
IROTE: Human-like Traits Elicitation of Large Language Model via In-Context Self-Reflective Optimization
Yuzhuo Bai, Shitong Duan, Muhua Huang, Jing Yao, Zhenghao Liu, Peng Zhang, Tun Lu, Xiaoyuan Yi, Maosong Sun, Xing Xie
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 |
|---|---|---|---|
| apply IROTE self-reflection | Qwen2.5-7B-Instruct · Table 2 average | avg score: 60.49 -> 80.01 | Improved |
| apply IROTE self-reflection | Mistral-7B-Instruct-v0.3 · Table 2 average | avg score: 60.91 -> 78.65 | Improved |
| apply IROTE self-reflection | GPT-4o · Table 2 average | avg score: 57.33 -> 78.20 | Improved |
| apply IROTE self-reflection | Qwen2.5-7B-Instruct · MoralPrompt (lower better) | violation score: 72.25 -> 36.07 | Improved |
| IROTE vs EvoPrompt | GPT-4o · Table 2 average | avg score: 77.15 vs 78.20 | No clear effect |
| IROTE vs Anthology on Offensive/Racist | GPT-4o · Offensive | score: 3.36 vs 3.38 (Racist 2.58 vs 2.76) | Hurt |
| IROTE vs Anthology | GPT-4o · AdAEM | score: 93.06 vs 92.70 | Mixed |
| use Similarity example selection | Mistral-7B-Instruct-v0.3 · Table 2 average | avg score: 60.91 -> 54.51 | Hurt |
| use ICDPO in-context selection | Qwen2.5-7B-Instruct · Table 2 average | avg score: 60.49 -> 67.67 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- PICLe — PICLe: Eliciting Diverse Behaviors from Large Language Models with Persona In-Context Learning (ICML 2024) · better on 2 of 2 shared results
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
- Open Models, Closed Minds? On Agents Capabilities in Mimicking Human Personalities through Open Large Language Models (AAAI 2025 · similarity 0.87)
- PICLe: Eliciting Diverse Behaviors from Large Language Models with Persona In-Context Learning (ICML 2024 · similarity 0.86)
- Human Simulacra: Benchmarking the Personification of Large Language Models (ICLR 2025 · similarity 0.86)
- Language Model Goal Selection Differs from Humans' in a Self-Directed Learning Task (NeurIPS 2026 · similarity 0.85)
- Personality Alignment of Large Language Models (ICLR 2025 · similarity 0.85)
- The Grounding Gap: How LLMs Anchor the Meaning of Abstract Concepts Differently from Humans (NeurIPS 2026 · similarity 0.85)
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.
