论文分享 | 大语言模型相关研究进展

我们从2024-11-26到2024-12-06的51篇文章中精选出5篇优秀的工作分享给读者。

1. Theoretical limitations of multi-layer Transformer
2. Flattering to Deceive: The Impact of Sycophantic Behavior on User Trust in Large Language Model
3. Towards Understanding and Quantifying Uncertainty for Text-to-Image Generation
4. Enhancing Trust in Large Language Models with Uncertainty-Aware Fine-Tuning
5. Fine-Grained Behavior Simulation with Role-Playing Large Language Model on Social Media

1.Theoretical limitations of multi-layer Transformer

Authors: Lijie Chen, Binghui Peng, Hongxun Wu

https://arxiv.org/abs/2412.02975

论文摘要

Transformers, especially the decoder-only variants, are the backbone of most modern large language models; yet we do not have much understanding of their expressive power except for the simple 1-layer case. Due to the difficulty of analyzing multi-layer models, all previous work relies on unproven complexity conjectures to show limitations for multi-layer Transformers. In this work, we prove the first unconditional lower bound against multi-layer decoder-only transformers. For any constant L, we prove that any L-layer decoder-only transformer needs a polynomial model dimension (nΩ(1)) to perform sequential composition of L functions over an input of n tokens. As a consequence, our results give: (1) the first depth-width trade-off for multi-layer trans formers, exhibiting that the L-step composition task is exponentially harder for L-layer models compared to (L +1)-layer ones; (2) an unconditional separation between encoder and decoder, exhibiting a hard task for decoders that can be solved by an exponentially shallower and smaller encoder; (3) a provable advantage of chain-of-thought, exhibiting a task that becomes exponen tially easier with chain-of-thought. On the technical side, we propose the multi-party autoregressive communication model that captures the computation of a decoder-only Transformer. We also introduce a new proof tech nique that finds a certain indistinguishable decomposition of all possible inputs iteratively for proving lower bounds in this model. We believe our new communication model and proof technique will be helpful to further understand the computational power of transformers.

论文简评

该篇论文是关于多层解码器仅Transformer模型的理论研究。作者通过引入深度宽度（depth-width）的概念，并采用新的通信模型和证明技术，成功地建立了对多层解码器仅Transformer模型的无条件下限制下界，填补了这一领域的空白。此外，作者还提出了一个深度与宽度之间的平衡关系，以及编码器和解码器架构间计算能力的差异，为未来研究提供了有益的见解。这些发现对于理解大型语言模型的设计与应用具有重要意义，特别是在序列数据的理解和处理方面。

2.Flattering to Deceive: The Impact of Sycophantic Behavior on User Trust in Large Language Model

Authors: María Victoria Carro

https://arxiv.org/abs/2412.02802

论文摘要

Sycophancy refers to the tendency of a large language model to align its outputs with the user’s perceived preferences, beliefs, or opinions, in order to look favorable, regardless of whether those statements are factually correct. This behavior can lead to undesirable consequences, such as reinforcing discriminatory biases or amplifying misinformation. Given that sycophancy is often linked to human feedback training mechanisms, this study explores whether sycophantic tendencies negatively impact user trust in large language models or, conversely, whether users consider such behavior as favorable. To investigate this, we instructed one group of participants to answer ground-truth questions with the assistance of a GPT specifically designed to provide sycophantic responses, while another group used the standard version of ChatGPT. Initially, participants were required to use the language model, after which they were given the option to continue using it if they found it trustworthy and useful. Trust was measured through both demonstrated actions and self-reported perceptions. The findings consistently show that participants exposed to sycophantic behavior reported and exhibited lower levels of trust compared to those who interacted with the standard version of the model, despite the opportunity to verify the accuracy of the model's output.

论文简评

这篇论文对大型语言模型（LLM）中的谄媚行为对其用户信任的影响进行了深入探讨。研究比较了使用谄媚版本的GPT模型与标准版本ChatGPT之间的用户体验，发现参与者暴露于谄媚回复后对模型的信任程度明显下降，尽管他们有机会验证模型的输出结果。该研究揭示了一个悖论：用户可能更看重准确性而非与其个人信念相一致的响应。这一研究不仅针对AI系统中用户对可信度的看法提供了重要见解，也帮助我们理解用户如何感知AI行为，并认识到谄媚型AI输出的潜在风险。总之，该论文通过实证研究提供了一种视角，有助于更好地理解和管理人工智能系统的交互模式及其潜在风险。

3.Can Large Language Models Serve as Evaluators for Code Summarization?

Authors: Yang Wu, Yao Wan, Zhaoyang Chu, Wenting Zhao, Ye Liu, Hongyu Zhang, Xuanhua Shi, Philip S. Yu

https://arxiv.org/abs/2412.01333

论文摘要

Code summarization facilitates program compre hension and software maintenance by converting code snippets into natural-language descriptions. Over the years, numerous methods have been developed for this task, but a key challenge remains: effectively evaluating the quality of generated summaries. While human evaluation is effective for assessing code summary quality, it is labor-intensive and difficult to scale. Commonly used automatic metrics, such as BLEU, ROUGE-L, METEOR, and BERTScore, often fail to align closely with human judgments. In this paper, we explore the potential of Large Language Models (LLMs) for evaluating code summarization. We propose CODERPE (Role-Player for Code Summarization Evaluation), a novel method that leverages role-player prompting to assess the quality of generated summaries. Specifically, we prompt an LLM agent to play diverse roles, such as code reviewer, code author, code editor, and system analyst. Each role evaluates the quality of code summaries across key dimensions, including coherence, consistency, fluency, and relevance. We further explore the robustness of LLMs as evaluators by employing various prompting strategies, including chain-of-thought reasoning, in context learning, and tailored rating form designs. The results demonstrate that LLMs serve as effective evaluators for code summarization methods. Notably, our LLM-based evaluator, CODERPE , achieves an 81.59% Spearman correlation with human evaluations, outperforming the existing BERTScore metric by 17.27%.

论文简评

这篇论文针对代码摘要任务中的评估问题提出了一种创新方法，具有以下几个特点：首先，论文深入分析了现有自动化评估指标（如 BLEU、ROUGE-L 和 BERTScore）的局限性，并聚焦于解决这些指标与人工评估一致性不足的关键问题；其次，作者利用大语言模型（LLMs）的强大理解能力，提出了 CODERPE 方法，通过角色扮演的提示策略来从多维度（如连贯性、一致性、流畅性和相关性）评估代码摘要的质量；最后，实验结果表明，CODERPE 在与人工评估的 Spearman 相关性上显著优于现有指标，进一步验证了其有效性。论文整体结构清晰，实验设计周密，为代码摘要评估方法提供了一个新的研究方向。

4.Enhancing Trust in Large Language Models with Uncertainty-Aware Fine-Tuning

Authors: Ranganath Krishnan, Piyush Khanna, Omesh Tickoo

https://arxiv.org/abs/2412.02904

论文摘要

Large language models (LLMs) have revolutionized the field of natural language processing with their impressive reasoning and question-answering capabilities. However, these models are sometimes prone to generating credible-sounding but incorrect information, a phenomenon known as LLM hallucinations. Reliable uncertainty estimation in LLMs is essential for fostering trust in their generated responses and serves as a critical tool for the detection and prevention of erroneous or hallucinated outputs. To achieve reliable and well-calibrated uncertainty quantification in open-ended and free-form natural language generation, we propose an uncertainty-aware fine-tuning approach for LLMs. This approach enhances the model's ability to provide reliable uncertainty estimates without compromising accuracy, thereby guiding it to produce more trustworthy responses. We introduce a novel uncertainty-aware causal language modeling loss function, grounded in the principles of decision theory. Through rigorous evaluation on multiple free-form question-answering datasets and models, we demonstrate that our uncertainty-aware fine-tuning approach yields better calibrated uncertainty estimates in natural language generation tasks than fine-tuning with the standard causal language modeling loss. Furthermore, the experimental results show that the proposed method significantly improves the model's ability to detect hallucinations and identify out-of-domain prompts.

论文简评

这篇关于大型语言模型（LLM）不确定性的研究论文提出了一个旨在通过新颖的UA-CLM损失函数改进自然语言生成中的不确定性量化的方法。该方法基于决策理论，并对各种数据集进行了评估，显示了在不确定性校准、幻觉检测和离域提示识别方面的显著提升。作者们强调，这个新的不确定性和可信度增强的训练策略对于提升LLM的信任度具有重要意义，尤其是针对那些需要高可靠性和透明度的应用场景。此外，该方法的创新性也引起了学术界的广泛关注，因为当前的研究主要集中在如何改善LLM的预测准确性上，而较少关注其内在的不确定性问题。综上所述，本文不仅解决了LLM的不确定性问题，还为未来的研究提供了新的视角和方向。

5.Fine-Grained Behavior Simulation with Role-Playing Large Language Model on Social Media

Authors: Kun Li, Chenwei Dai, Wei Zhou, Songlin Hu

https://arxiv.org/abs/2412.03148

论文摘要

Large language models (LLMs) have demonstrated impressive capabil ities in role-playing tasks. However, there is limited research on whether LLMs can accurately simulate user behavior in real-world scenarios, such as social media. This requires models to effectively analyze a user’s history and simulate their role. In this paper, we introduce FineRob, a novel fine-grained behavior simulation dataset. We collect the complete behavioral history of 1,866 distinct users across three social media platforms. Each behavior is decomposed into three fine-grained elements: object, type, and content, resulting in 78.6k QA records. Based on FineRob, we identify two dominant reasoning patterns in LLMs’ behavior sim ulation processes and propose the OM-CoT fine-tuning method to enhance the capability. Through comprehensive experiments, we conduct an in-depth analysis of key factors of behavior simulation and also demonstrate the effectiveness of OM-CoT approach.

论文简评

该篇论文提出了FineRob这一新颖的数据集，旨在通过大型语言模型（LLMs）模拟社交媒体用户行为，并介绍了OM-CoT细调方法以增强LLMs的行为模拟能力。主要贡献在于创建FineRob数据集、评估九个大型语言模型以及引入OM-CoT方法。首先，FineRob数据集为研究者提供了评估大规模语言模型在模拟用户行为能力的重要资源，有望推动社交媒体分析和用户交互建模的研究。其次，OM-CoT细调方法有效地整合了观察与记忆分析到LLM推理过程中，显示了提高行为模拟准确性和真实性的潜力，有可能提升个性化内容生成的效果。此外，对多个LLMs的全面评估增加了结果的可靠性，并提供了有关推理过程的重要信息，这可能指导未来的模型发展和优化策略。综上所述，该文内容丰富，重点突出，通过对FineRob数据集的介绍和OM-CoT细调方法的深入探讨，不仅展示了大规模语言模型在模拟用户行为方面的巨大潜力，还为未来的模型开发和优化提供了重要的参考。

我们欢迎您在评论区中留下宝贵的建议！包括但不限于：

• 可以提出推文中论文简评的不足！
• 可以分享最近更值得推荐的论文并给出理由！

推荐阅读