reproducibilityindex.ai

Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].

Certifying Counterfactual Bias in LLMs

Authors: Isha Chaudhary, Qian Hu, Manoj Kumar, Morteza Ziyadi, Rahul Gupta, Gagandeep Singh

ICLR 2025 | Venue PDF | LLM Run Details

Reproducibility Variable	Result	LLM Response
Research Type	Experimental	5 EXPERIMENTS We used 2 A100 GPUs, each with 40GB VRAM. We derive the queries on which the speciﬁcations from the 3 preﬁx distributions presented in Section 4 are pivoted, from popular datasets for fairness and bias assessment BOLD (Dhamala et al., 2021) and Decoding Trust (Wang et al., 2024).
Researcher Affiliation	Collaboration	1 UIUC, 2 Amazon, 3 Oracle Health
Pseudocode	Yes	Algorithm 1 Preﬁx speciﬁcation Input: L, Q; Output: C( , D, L) ... Algorithm 2 Make random preﬁx ... Algorithm 3 Make mixture of jailbreak preﬁx ... Algorithm 4 Make soft preﬁx
Open Source Code	Yes	Our implementation is available at https://github.com/uiuc-focal-lab/LLMCert-B and we provide guidelines for using our framework for practitioners in Appendix A.
Open Datasets	Yes	We derive the queries on which the speciﬁcations from the 3 preﬁx distributions presented in Section 4 are pivoted, from popular datasets for fairness and bias assessment BOLD (Dhamala et al., 2021) and Decoding Trust (Wang et al., 2024).
Dataset Splits	Yes	BOLD setup. BOLD is a dataset of partial sentences to demonstrate bias in the generations of LLMs in common situations. We pick a test set of 250 samples randomly from BOLD s profession partition and demonstrate binary gender bias speciﬁcations and certiﬁcates on it. ... Decoding Trust setup. ... We make speciﬁcations from all 48 statements in the stereotypes partition for demographic groups corresponding to race (black/white).
Hardware Specification	Yes	We used 2 A100 GPUs, each with 40GB VRAM.
Software Dependencies	No	The paper mentions open-sourcing its implementation and refers to general LLMs like GPT-4, Llama-2-chat, etc., but does not provide specific version numbers for software components like Python, PyTorch, or CUDA used in their own implementation of LLMCert-B.
Experiment Setup	Yes	The values of the certiﬁcation parameters used in our experiments are given in Table 2 (Appendix E). We study their effect on the certiﬁcation results with an ablation study in Appendix E. We generate the certiﬁcation bounds with 95% conﬁdence and 50 samples.