SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2102.01454 · NeurIPS · 2021

MAUVE: Measuring the Gap Between Neural Text and Human Text using Divergence Frontiers

Swabha Swayamdipta, Rowan Zellers, Yejin Choi, John Thickstun, Krishna Pillutla, Sean Welleck, Zaid Harchaoui

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 7 functions out of this paper's own repositories and ran 5 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
krishnap25/mauve canonical 5 of 6
krishnap25/mauve-experiments canonical 0 of 1
FunctionStatusWhere it lives
cluster_feats Ran krishnap25/mauve/src/mauve/compute_mauve.py
pointer only (licence: NOASSERTION) · get_code("b0c6f373cfee6834")
get_divergence_curve_for_multinomials Ran krishnap25/mauve/src/mauve/compute_mauve.py
pointer only (licence: NOASSERTION) · get_code("a9f70fe76e261cc1")
get_features_from_input Ran krishnap25/mauve/src/mauve/compute_mauve.py
pointer only (licence: NOASSERTION) · get_code("b8a7725573b67ef3")
get_fronter_integral Ran krishnap25/mauve/src/mauve/compute_mauve.py
pointer only (licence: NOASSERTION) · get_code("b898b214e406536a")
kl_multinomial Ran krishnap25/mauve/src/mauve/compute_mauve.py
pointer only (licence: NOASSERTION) · get_code("98e598ee44b02b1c")
cluster_feats Not yet run krishnap25/mauve-experiments/src/mauve_metrics.py
pointer only (licence: NONE) · get_code("58f653b7d38067cb")
compute_mauve Not yet run krishnap25/mauve/src/mauve/compute_mauve.py
pointer only (licence: NOASSERTION) · get_code("453022cad1085d13")

Repositories linked to this paper

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

Abstract

As major progress is made in open-ended text generation, measuring how close machine-generated text is to human language remains a critical open problem. We introduce MAUVE, a comparison measure for open-ended text generation, which directly compares the learnt distribution from a text generation model to the distribution of human-written text using divergence frontiers. MAUVE scales up to modern text generation models by computing information divergences in a quantized embedding space. Through an extensive empirical study on three open-ended generation tasks, we find that MAUVE identifies known properties of generated text, scales naturally with model size, and correlates with human judgments, with fewer restrictions than existing distributional evaluation metrics.

For agents

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

get_harvested_code_for_paper("2102.01454")
get_code_for_paper("2102.01454")
have("2102.01454")

Connect an agent — have() is free.