# The Oracle Has Spoken: A Multi-Aspect Evaluation of Dialogue in Pythia

arXiv: https://arxiv.org/abs/2509.16487 · HTML: https://mlsanity.com/paper/2509.16487/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 10 of its 18 references that are in the map.
- First posted on arXiv 20 Sep 2025.

## Builds on (most cited first, 10 of 10)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [Pythia: A Suite for Analyzing Large Language Models Across Training and Scaling](https://mlsanity.com/paper/2304.01373/) (2023)
- [LIMA: Less Is More for Alignment](https://mlsanity.com/paper/2305.11206/) (2023)
- [PyTorch FSDP: Experiences on Scaling Fully Sharded Data Parallel](https://mlsanity.com/paper/2304.11277/) (2023)
- [Towards a Unified Multi-Dimensional Evaluator for Text Generation](https://mlsanity.com/paper/2210.07197/) (2022)
- [The False Promise of Imitating Proprietary LLMs](https://mlsanity.com/paper/2305.15717/) (2023)
- [Themis: A Reference-free NLG Evaluation Language Model with Flexibility and Interpretability](https://mlsanity.com/paper/2406.18365/) (2024)
- [Should We Fine-Tune or RAG? Evaluating Different Techniques to Adapt LLMs for Dialogue](https://mlsanity.com/paper/2406.06399/) (2024)

## Abstract (arXiv)

Dialogue is one of the landmark abilities of large language models (LLMs). Despite its ubiquity, few studies actually distinguish specific ingredients underpinning dialogue behavior emerging during post-training. We employ a comprehensive suite of model-based metrics, each targeting a distinct fine-grained aspect of dialogue, motivated by linguistic theory. We evaluate how the performance of pre-trained Pythia models changes with respect to each of those dimensions, depending on model size and as a result of supervised fine-tuning on conversational datasets. We observe only a mild impact of raw model size on most metrics, whereas fine-tuning quickly saturates the scores for all but the smallest models tested. Somewhat contrary to our expectations, many metrics show very similar trends, especially if they are all rooted in the same evaluator model, which raises the question of their reliability in measuring a specific dimension. To that end, we conduct additional analyses of score distributions, metric correlations, and term frequencies in generated responses to help explain our observations.