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].
More the Merrier: Towards Multi-Emotion and Intensity Controllable Response Generation
Authors: Mauajama Firdaus, Hardik Chauhan, Asif Ekbal, Pushpak Bhattacharyya12821-12829
AAAI 2021 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | The detailed evaluation shows that our proposed approach attains superior performance compared to the baseline models. Experiments Implementation details For all the models including baselines, the batch size is set to 32. The utterance encoder is a bidirectional GRU with 600 hidden units in each direction. |
| Researcher Affiliation | Academia | Mauajama Firdaus, Hardik Chauhan, Asif Ekbal and Pushpak Bhattacharyya Department of Computer Science and Engineering Indian Institute of Technology Patna, India (mauajama.pcs16,hardik,asif,pb)@iitp.ac.in |
| Pseudocode | No | No structured pseudocode or algorithm blocks were found. |
| Open Source Code | No | No explicit statement or link providing access to the source code for the methodology was found. |
| Open Datasets | No | Due to the unavailability of multi-emotion intensity labeled data for our proposed task, we create a large-scale dialogue dataset, MEIMD from 8 English TV series having 34k conversations that have been labeled with multiple emotions and their intensities. We create a large-scale Multiple Emotion and Intensity aware Multi-party Dialogue (MEIMD) dataset. |
| Dataset Splits | No | The paper does not explicitly provide training/test/validation dataset splits needed to reproduce the experiment. |
| Hardware Specification | No | No specific hardware details (GPU/CPU models, processor types, memory amounts, or detailed computer specifications) used for running experiments were found. |
| Software Dependencies | No | The paper mentions optimizers and embeddings (AMS-Grad, Glove) but does not provide specific version numbers for software dependencies or libraries. |
| Experiment Setup | Yes | For all the models including baselines, the batch size is set to 32. The utterance encoder is a bidirectional GRU with 600 hidden units in each direction. The context encoder and decoder are both GRUs with 600 hidden units. All the model parameters are randomly initialised using a Gaussian distribution with Xavier scheme (Glorot and Bengio 2010). We use 0.45 as dropout rate and perform gradient clipping when gradient norm is over 3. |