Capturing the Relationship Between Sentence Triplets for LLM and Human-Generated Texts to Enhance Sentence Embeddings
Bibliographic record
Abstract
Deriving meaningful sentence embeddings is crucial in capturing the semantic relationship between texts.Recent advances in building sentence embedding models have centered on replacing traditional human-generated text datasets with those generated by LLMs.However, the properties of these widely used LLMgenerated texts remain largely unexplored.Here, we evaluate the quality of the LLMgenerated texts from four perspectives (Positive Text Repetition, Length Difference Penalty, Positive Score Compactness, and Negative Text Implausibility) and find the limitation of only using LLM to build high-quality NLI datasets.Then, we attempt to improve each of these models either fine-tuned with human, LLM, or human+LLM-generated sentence triplets data with our proposed loss function that incorporates Positive-Negative sample Augmentation (PNA) within the contrastive learning objective.Our results demonstrate the effectiveness of PNA, especially in RoBERTalarge, by showing decreased cosine similarity for sentence triplets, mitigating the sentence anisotropy problem in Wikipedia corpus (-7% compared to CLHAIF), and improving the Spearman's correlation in standard Semantic Textual Similarity (STS) tasks (+1.47% compared to CLHAIF).Our code is available at https://github.com/xfactlab/eacl2024-pna. SimCSE
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.012 |
| Meta-epidemiology (narrow) | 0.002 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.006 | 0.005 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".