Collaborative Alignment of NLP Models

Authors: Fereshte Khani, Marco Tulio Ribeiro

NeurIPS 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Our experiments show Co Align is effective at helping multiple users operationalize concepts and avoid interference for a variety of scenarios, tasks, and models.
Researcher Affiliation Industry Fereshte Khani Microsoft fkhani@microsoft.com Marco Tulio Ribeiro Google Deep Mind marcotcr@gmail.com
Pseudocode Yes Algorithm 1: Operationalizing a new concept i
Open Source Code No Code and data will be released in https://github.com/fereshte-khani/Co Align.
Open Datasets Yes We use Quora Question Pairs (QQP) dataset... We train RoBERTa-Large on the whole Amazon Review dataset... We use RoBERTa-Large [18] finetuned on MNLI (binary) [19]...
Dataset Splits No The paper mentions 'base validation dataset' for the MNLI task. However, for all experiments, it does not provide explicit overall training/validation/test split percentages, absolute sample counts, or detailed splitting methodology needed for full reproduction of data partitioning.
Hardware Specification No The paper does not provide any specific details about the hardware (e.g., GPU model, CPU type, memory) used for conducting the experiments.
Software Dependencies No The paper mentions using GPT-3 and fine-tuning RoBERTa-Large, but it does not specify version numbers for these models or any other software dependencies, libraries, or frameworks used in the implementation.
Experiment Setup No While the paper describes various experimental tasks and model evaluations, it does not explicitly provide concrete hyperparameter values (e.g., learning rate, batch size, number of epochs) or detailed system-level training configurations for reproducibility of the experimental setup.