Research map / Large language models / Alignment and preferences
Verification and Co-Alignment via Heterogeneous Consistency for Preference-Aligned LLM Annotations
Cheng Chen, Haiyan Yin, Ivor Tsang
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 HCC co-alignment | Meta-Llama-3-8B-Instruct · Clinc | accuracy (%): 32.49 -> 82.43 | Improved |
| apply HCC co-alignment | Meta-Llama-3-8B-Instruct · Massive Scenario | accuracy (%): 43.52 -> 78.13 | Improved |
| apply HCC co-alignment | GPT-4o Mini · Banking77 | accuracy (%): 65.12 -> 82.45 | Improved |
| apply HCC co-alignment | GPT-4o Mini · Clinc | accuracy (%): 81.44 -> 87.93 | Improved |
| apply HCC co-alignment | GPT-4o Mini · MTOP Intent | accuracy (%): 75.03 -> 67.10 | Hurt |
| apply HCC co-alignment | GPT-4o Mini · StackExchange | accuracy (%): 51.90 -> 45.22 | Hurt |
| add the correction (co-alignment) step on top of HCC without correction (ablation) | HCC with GPT-3.5 Turbo · MTOP Intent | accuracy (%): 79.57 -> 69.06 | Hurt |
| add the correction (co-alignment) step on top of HCC without correction (ablation) | HCC with GPT-4o Mini · MTOP Intent | accuracy (%): 80.16 -> 67.10 | Hurt |
| add the correction (co-alignment) step on top of HCC without correction (ablation) | HCC with GPT-3.5 Turbo · Clinc | accuracy (%): 81.32 -> 85.49 | Improved |
| use stronger MiniLM instead of BERT-Base as the specialised model (ablation) | HCC with Llama-3-8B Instruct · Banking77 | accuracy (%): 71.62 -> 77.71 | Improved |
| use stronger MiniLM instead of BERT-Base as the specialised model (ablation) | HCC with Llama-3-8B Instruct · Massive Scenario | accuracy (%): 74.61 -> 78.13 | Improved |
| use imbalanced user-preference samples (60% imbalance) instead of balanced (ablation) | HCC with Llama-3-8B Instruct · Clinc | accuracy drop: -0.32 | Hurt |
Closest papers by meaning
- Spread Preference Annotation: Direct Preference Judgment for Efficient LLM Alignment (ICLR 2025 · similarity 0.92)
- Stackelberg Self-Annotation: A Robust Approach to Data-Efficient LLM Alignment (NeurIPS 2025 · similarity 0.91)
- TODO: Enhancing LLM Alignment with Ternary Preferences (ICLR 2025 · similarity 0.91)
- Exploring Intrinsic Alignments Within Text Corpus (AAAI 2025 · similarity 0.90)
- MM-RLHF: The Next Step Forward in Multimodal LLM Alignment (ICML 2025 · similarity 0.90)
- Aligner: Efficient Alignment by Learning to Correct (NeurIPS 2024 · similarity 0.90)
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.
