Mitigating the Learning Bias towards Repetition by Self-Contrastive Training for Open-Ended Generation (2023.findings-acl)
Copied to clipboard
| Challenge: | Existing language models generate repetitive texts with greedy decoding or beam search. |
| Approach: | They propose a self-contrastive training technique to penalize the output of a premature checkpoint of the same model when it incorrectly predicts repetition. |
| Outcome: | The proposed training mitigates repetition while maintaining fluency while minimizing the overestimation of token-level repetition probabilities. |
Similar Papers
Contrastive Decoding: Open-ended Text Generation as Optimization (2023.acl-long)
Copied to clipboard
Xiang Lisa Li, Ari Holtzman, Daniel Fried, Percy Liang, Jason Eisner, Tatsunori Hashimoto, Luke Zettlemoyer, Mike Lewis
| Challenge: | Using a language model, maximum probability is a poor decoding objective because it produces short and repetitive text. |
| Approach: | They propose a reliable decoding approach that optimizes a contrastive objective subject to a plausibility constraint. |
| Outcome: | The proposed approach outperforms four strong decoding algorithms in automatic and human evaluations across wikipedia, news and story domains. |
Penalty Decoding: Well Suppress the Self-Reinforcement Effect in Open-Ended Text Generation (2023.emnlp-main)
Copied to clipboard
| Challenge: | Experimental results demonstrate the efficacy of our approach in generating high-quality sentences resembling human output. |
| Approach: | They propose a forgetting mechanism that disregards distant tokens, reducing the burden of penalty selection. |
| Outcome: | The proposed approach generates high-quality sentences resembling human output. |
Sequence-level Large Language Model Training with Contrastive Preference Optimization (2025.findings-naacl)
Copied to clipboard
| Challenge: | a new method to improve the performance of large language models requires a small computational cost. |
| Approach: | They propose a CPO procedure that can inject sequence-level information into the model at any training stage without expensive human labeled data. |
| Outcome: | The proposed objective surpasses the next token prediction in terms of win rate in instruction-following and text generation tasks. |
FOCUS: Forging Originality through Contrastive Use in Self-Plagiarism for Language Models (2024.findings-acl)
Copied to clipboard
| Challenge: | Existing methods to generate original text using pre-trained language models are problematic as they are trained on corpora constructed by human authors. |
| Approach: | They propose a unique “self-plagiarism” contrastive decoding strategy that modifies prompts in LLMs to develop an amateur model and a professional model. |
| Outcome: | The proposed method enables the development of an amateur model and a professional model while maintaining its standard language model status. |
Rethinking Repetition Problems of LLMs in Code Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies have focused on content repetition, but structural repetition is a more prevalent problem in code generation. |
| Approach: | They propose a decoding approach that eliminates repetition problems in code generation by identifying grammar rules and strategically decaying the likelihood of critical tokens that contribute to repetitions. |
| Outcome: | The proposed approach outperforms baselines and humanEval benchmarks on CodeRepetEval dataset and MBPP benchmarks, effectively reducing repetitions and enhancing the quality of generated code. |
Observations and Remedies for Large Language Model Bias in Self-Consuming Performative Loop (2026.acl-long)
Copied to clipboard
| Challenge: | Existing synthetic training loops for large language models cause performance drops and induce emerging biases . a large amount of generated content is posted to coding platforms, social media platforms and other platforms on the internet . |
| Approach: | They propose a self-consuming retraining loop where models are trained on their own outputs . they use a control loop to isolate and analyze feedback-driven bias evolution . |
| Outcome: | The proposed model increases preference bias and decreases disparate bias. |
MiLe Loss: a New Loss for Mitigating the Bias of Learning Difficulties in Generative Language Models (2024.findings-naacl)
Copied to clipboard
| Challenge: | Existing generative language models neglect an inherent challenge in text corpus during training, i.e., the imbalance between frequent tokens and infrequent ones. |
| Approach: | They propose a function to mitigate the imbalance between frequent and infrequent tokens . authors propose 'MiLe Loss' function to assess learning difficulty of tokens during training . |
| Outcome: | Experiments show that models with proposed model can improve on downstream benchmarks. |
Implicit Unlikelihood Training: Improving Neural Text Generation with Reinforcement Learning (2021.eacl-main)
Copied to clipboard
| Challenge: | Existing approaches to language modeling use autoregressive methods, but they can produce repetitive results. |
| Approach: | They propose to add a loss function for regularization to avoid unwanted properties, such as contradiction or repetition, to a language model by using policy gradient reinforcement learning. |
| Outcome: | The proposed method reduces repetition without impacting the language model quality. |
Generation-driven Contrastive Self-training for Zero-shot Text Classification with Instruction-following LLM (2024.eacl-long)
Copied to clipboard
| Challenge: | a novel method to train a smaller model with LLMs for zero-shot text classification requires immense computational resources due to their substantial model size. |
| Approach: | They propose a method which leverages the generative power of large language models to train a smaller model. |
| Outcome: | The proposed method outperforms state-of-the-art methods when limited data is available. |
Enhancing Contextual Understanding in Large Language Models through Contrastive Decoding (2024.naacl-long)
Copied to clipboard
| Challenge: | Large language models lack contextual knowledge, resulting in text with factual inconsistencies or contextually unfaithful content. |
| Approach: | They propose a method that integrates contrastive decoding with adversarial irrelevant passages as negative samples to enhance robust context grounding during generation. |
| Outcome: | The proposed method improves context grounding during generation without training. |