SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2212.07919 · 2022

ROSCOE: A Suite of Metrics for Scoring Step-by-Step Reasoning

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 9 functions out of this paper's own repositories and ran 8 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.

RepositoryRoleRan
facebookresearch/ParlAI canonical 8 of 9
FunctionStatusWhere it lives
escape Ran facebookresearch/ParlAI/parlai/core/dict.py
pointer only (licence: MIT) · get_code("758e9948df81569f")
find_ngrams Ran facebookresearch/ParlAI/parlai/core/dict.py
pointer only (licence: MIT) · get_code("761ab53c3cc0c21a")
get_metric_display_data Ran facebookresearch/ParlAI/parlai/core/metrics.py
pointer only (licence: MIT) · get_code("b6eb8b2b6abaf6ba")
register_agent Ran facebookresearch/ParlAI/parlai/core/loader.py
pointer only (licence: MIT) · get_code("60339a7726e86b18")
register_mutator Ran facebookresearch/ParlAI/parlai/core/mutators.py
pointer only (licence: MIT) · get_code("f9ccfd237025f9b0")
register_script Ran facebookresearch/ParlAI/parlai/core/loader.py
pointer only (licence: MIT) · get_code("eba5f047c4021087")
register_teacher Ran facebookresearch/ParlAI/parlai/core/loader.py
pointer only (licence: MIT) · get_code("b965ad8e264b80f1")
unescape Ran facebookresearch/ParlAI/parlai/core/dict.py
pointer only (licence: MIT) · get_code("7e5138ac467c5625")
normalize_answer Not yet run facebookresearch/ParlAI/parlai/core/metrics.py
pointer only (licence: MIT) · get_code("da3bc8c42eec18d1")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Large language models show improved downstream task performance when prompted to generate step-by-step reasoning to justify their final answers. These reasoning steps greatly improve model interpretability and verification, but objectively studying their correctness (independent of the final answer) is difficult without reliable methods for automatic evaluation. We simply do not know how often the stated reasoning steps actually support the final end task predictions. In this work, we present ROSCOE, a suite of interpretable, unsupervised automatic scores that improve and extend previous text generation evaluation metrics. To evaluate ROSCOE against baseline metrics, we design a typology of reasoning errors and collect synthetic and human evaluation scores on commonly used reasoning datasets. In contrast with existing metrics, ROSCOE can measure semantic consistency, logicality, informativeness, fluency, and factuality - among other traits - by leveraging properties of step-by-step rationales. We empirically verify the strength of our metrics on five human annotated and six programmatically perturbed diagnostics datasets - covering a diverse set of tasks that require reasoning skills and show that ROSCOE can consistently outperform baseline metrics.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2212.07919")
get_code_for_paper("2212.07919")
have("2212.07919")

Connect an agent — have() is free.