Pith. sign in

REVIEW 6 cited by

Want To Reduce Labeling Cost? GPT-3 Can Help

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2108.13487 v1 pith:437CQS27 submitted 2021-08-30 cs.CL cs.AI

Want To Reduce Labeling Cost? GPT-3 Can Help

Shuohang Wang , Yang Liu , Yichong Xu , Chenguang Zhu , Michael Zeng This is my paper
classification cs.CL cs.AI
keywords datagpt-3labelslabelingmanytasksmodelperformance
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Data annotation is a time-consuming and labor-intensive process for many NLP tasks. Although there exist various methods to produce pseudo data labels, they are often task-specific and require a decent amount of labeled data to start with. Recently, the immense language model GPT-3 with 175 billion parameters has achieved tremendous improvement across many few-shot learning tasks. In this paper, we explore ways to leverage GPT-3 as a low-cost data labeler to train other models. We find that, to make the downstream model achieve the same performance on a variety of NLU and NLG tasks, it costs 50% to 96% less to use labels from GPT-3 than using labels from humans. Furthermore, we propose a novel framework of combining pseudo labels from GPT-3 with human labels, which leads to even better performance with limited labeling budget. These results present a cost-effective data labeling methodology that is generalizable to many practical applications.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 6 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Dynamic Emotion and Personality Profiling for Multimodal Deception Detection

    cs.CL 2026-04 unverdicted novelty 6.0

    A new dataset DDEP and reliability-weighted fusion model Rel-DDEP jointly detect deception, emotion, and personality from multimodal data, reporting F1 gains of 2.53%, 2.66%, and 9.30% over baselines.

  2. Distilling Step-by-Step! Outperforming Larger Language Models with Less Training Data and Smaller Model Sizes

    cs.CL 2023-05 conditional novelty 6.0

    Distilling step-by-step uses LLM-generated rationales as additional supervision in a multi-task framework so that 770M-parameter models outperform 540B-parameter models on NLP benchmarks with only 80% of the data.

  3. Characterizing initial human-AI proof formalization workflows

    cs.AI 2026-06 unverdicted novelty 5.0

    A controlled user study and qualitative survey find that AI assistance raises formalization accuracy for math proofs, with users flexibly combining multiple tools while retaining oversight.

  4. Structured Exploration and Exploitation of Label Functions for Automated Data Annotation

    cs.LG 2026-03 unverdicted novelty 5.0

    EXPONA improves automated data labeling by exploring multi-level label functions and applying reliability filters, achieving up to 98.9% coverage and 46% gains in downstream weighted F1 on eleven datasets.

  5. Learning to Select Visual In-Context Demonstrations

    cs.LG 2026-03 reject novelty 5.0

    A Dueling-DQN agent selects visual in-context demonstrations and outperforms kNN retrieval on objective regression benchmarks but not on subjective preference tasks, per the paper's main table.

  6. Refining and Reusing Annotation Guidelines for LLM Annotation

    cs.CL 2026-05 conditional novelty 4.0

    An iterative moderation framework refines and reuses annotation guidelines to improve LLM annotation accuracy on biomedical NER tasks across GPT, Gemini, and DeepSeek models.