We lifted 3 functions out of this paper's own repositories and ran 2 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.
| Repository | Role | Ran |
|---|---|---|
| ncsoft/offsetbias | canonical | 2 of 3 |
| Function | Status | Where it lives |
|---|---|---|
| get_model_statistics | Ran | ncsoft/offsetbias/run_bench.py code served (permissive licence) · get_code("34978fa5a31a0e95") |
| make_data_row | Ran | ncsoft/offsetbias/run_bench.py code served (permissive licence) · get_code("2c4b0375a2ce745c") |
| get_benchmark_data | Not yet run | ncsoft/offsetbias/run_bench.py code served (permissive licence) · get_code("54252129bd845012") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
Employing Large Language Models (LLMs) to assess the quality of generated responses, such as prompting instruct-tuned models or fine-tuning judge models, has become a widely adopted evaluation method. It is also known that such evaluators are vulnerable to biases, such as favoring longer responses. While it is important to overcome this problem, the specifics of these biases remain under-explored. In this work, we qualitatively identify six types of biases inherent in various judge models. We propose EvalBiasBench as a meta-evaluation collection of hand-crafted test cases for each bias type. Additionally, we present de-biasing dataset construction methods and the associated preference dataset OffsetBias. Experimental results demonstrate that fine-tuning on our dataset significantly enhances the robustness of judge models against biases and improves performance across most evaluation scenarios. We release our datasets and the fine-tuned judge model to public.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2407.06551")
get_code_for_paper("2407.06551")
have("2407.06551")
Connect an agent — have() is free.