GAPO: Learning Preferential Prompt through Generative Adversarial Policy Optimization

Recent advances in large language models have highlighted the critical need for precise control over model outputs through predefined constraints. While existing methods attempt to achieve this through either direct instruction-response synthesis or preferential response optimization, they often struggle with constraint understanding and adaptation. This limitation becomes particularly evident when handling fine-grained constraints, leading to either hallucination or brittle performance. We introduce Generative Adversarial Policy Optimization (GAPO), a novel framework that combines GAN-based training dynamics with an encoder-only reward model to progressively learn and adapt to increasingly complex constraints. GAPO leverages adversarial training to automatically generate training samples of varying difficulty while utilizing the encoder-only architecture to better capture prompt-response relationships. Extensive experiments demonstrate GAPO's superior performance across multiple benchmarks, particularly in scenarios requiring fine-grained constraint handling, where it significantly outperforms existing methods like PPO, DPO, and KTO. Our results suggest that GAPO's unique approach to preferential prompt learning offers a more robust and effective solution for controlling LLM outputs. Code is avaliable in https://github.com/MikeGu721/GAPO.

LLM-GAN: Construct Generative Adversarial Network Through Large Language Models For Explainable Fake News Detection

Explainable fake news detection predicts the authenticity of news items with annotated explanations. Today, Large Language Models (LLMs) are known for their powerful natural language understanding and explanation generation abilities. However, presenting LLMs for explainable fake news detection remains two main challenges. Firstly, fake news appears reasonable and could easily mislead LLMs, leaving them unable to understand the complex news-faking process. Secondly, utilizing LLMs for this task would generate both correct and incorrect explanations, which necessitates abundant labor in the loop. In this paper, we propose LLM-GAN, a novel framework that utilizes prompting mechanisms to enable an LLM to become Generator and Detector and for realistic fake news generation and detection. Our results demonstrate LLM-GAN's effectiveness in both prediction performance and explanation quality. We further showcase the integration of LLM-GAN to a cloud-native AI platform to provide better fake news detection service in the cloud.

StrucText-Eval: An Autogenerated Benchmark for Evaluating Large Language Model's Ability in Structure-Rich Text Understanding

Given the substantial volumes of structured data held by many companies, enabling Large Language Models (LLMs) to directly understand structured text in non-structured forms could significantly enhance their capabilities across various business scenarios. To this end, we propose evaluation data generation method for assessing LLM's ability in understanding the structure-rich text, which generates structured data of controllable complexity based on manually crafted question templates and generation rules. Building on this generation method, we introduce StrucText-Eval, a benchmark comprising 6,032 questions across 8 different structured languages and 29 specific tasks. Furthermore, considering human proficiency in rule-based tasks, we also present StrucText-Eval-Hard, which includes 3,016 questions designed to further examine the gap between LLMs and human performance. Results indicate that the best-performing LLM currently achieve an accuracy of 65.0\% on StrucText-Eval-Hard, while human accuracy reaches up to 95.7\%. Moreover, while fine-tuning using StrucText-Eval can enhance existing LLMs' understanding of all structured languages, it does not necessarily improve performance across all task types. The benchmark and generation codes are open sourced in https://github.com/MikeGu721/StrucText-Eval

DetectBench: Can Large Language Model Detect and Piece Together Implicit Evidence?

Detecting evidence within the context is a key step in the process of reasoning task. Evaluating and enhancing the capabilities of LLMs in evidence detection will strengthen context-based reasoning performance. This paper proposes a benchmark called DetectBench for verifying the ability to detect and piece together implicit evidence within a long context. DetectBench contains 3,928 multiple-choice questions, with an average of 994 tokens per question. Each question contains an average of 4.55 pieces of implicit evidence, and solving the problem typically requires 7.62 logical jumps to find the correct answer. To enhance the performance of LLMs in evidence detection, this paper proposes Detective Reasoning Prompt and Finetune. Experiments demonstrate that the existing LLMs' abilities to detect evidence in long contexts are far inferior to humans. However, the Detective Reasoning Prompt effectively enhances the capability of powerful LLMs in evidence detection, while the Finetuning method shows significant effects in enhancing the performance of weaker LLMs. Moreover, when the abilities of LLMs in evidence detection are improved, their final reasoning performance is also enhanced accordingly.

StructBench: An Autogenerated Benchmark for Evaluating Large Language Model's Ability in Structure-Rich Text Understanding

Given the substantial volumes of structured data held by many companies, enabling Large Language Models (LLMs) to directly understand structured text in non-structured forms could significantly enhance their capabilities across various business scenarios. To this end, we propose evaluation data generation method for assessing LLM's ability in understanding the structure-rich text, which generates structured data of controllable complexity based on manually crafted question templates and generation rules. Building on this generation method, we introduce StructBench, a benchmark comprising 6,032 questions across 8 different structured languages and 29 specific tasks. Furthermore, considering human proficiency in rule-based tasks, we also present StructBench-Hard, which includes 3,016 questions designed to further examine the gap between LLMs and human performance. Results indicate that the best-performing LLM currently achieve an accuracy of 65.0\% on StructBench-Hard, while human accuracy reaches up to 95.7\%. Moreover, while fine-tuning using StructBench can enhance existing LLMs' understanding of all structured languages, it does not necessarily improve performance across all task types. The benchmark and generation codes are open sourced in https://github.com/MikeGu721/StructBench

Agent Group Chat: An Interactive Group Chat Simulacra For Better Eliciting Collective Emergent Behavior

To investigate the role of language in human collective behaviors, we developed the Agent Group Chat simulation to simulate linguistic interactions among multi-agent in different settings. Agents are asked to free chat in this simulation for their own purposes based on their character setting, aiming to see agents exhibit emergent behaviours that are both unforeseen and significant. Four narrative scenarios, Inheritance Disputes, Law Court Debates, Philosophical Discourses, Movie Casting Contention, are integrated into Agent Group Chat to evaluate its support for diverse storylines. By configuring specific environmental settings within Agent Group Chat, we are able to assess whether agents exhibit behaviors that align with human expectations. We evaluate the disorder within the environment by computing the n-gram Shannon entropy of all the content speak by characters. Our findings reveal that under the premise of agents possessing substantial alignment with human expectations, facilitating more extensive information exchange within the simulation ensures greater orderliness amidst diversity, which leads to the emergence of more unexpected and meaningful emergent behaviors. The code is open source in https://github.com/MikeGu721/AgentGroup, and online platform will be open soon.

The Missing Piece in Model Editing: A Deep Dive into the Hidden Damage Brought By Model Editing

Large Language Models have revolutionized numerous tasks with their remarkable efficacy.However, the editing of these models, crucial for rectifying outdated or erroneous information, often leads to a complex issue known as the ripple effect in the hidden space. This effect, while difficult to detect, can significantly impede the efficacy of model editing tasks and deteriorate model performance.This paper addresses this scientific challenge by proposing a novel evaluation methodology, Graphical Outlier Relation based Assessment(GORA), which quantitatively evaluates the adaptations of the model and the subsequent impact of editing. Furthermore, we introduce the Selective Outlier Re-Editing Approach(SORA), a model editing method designed to mitigate this ripple effect. Our comprehensive evaluations reveal that the ripple effect in the hidden space is a significant issue in all current model editing methods. However, our proposed methods, GORA and SORA, effectively identify and alleviate this issue, respectively, contributing to the advancement of LLM editing techniques.

Beyond the Obvious: Evaluating the Reasoning Ability In Real-life Scenarios of Language Models on Life Scapes Reasoning Benchmark~(LSR-Benchmark)

This paper introduces the Life Scapes Reasoning Benchmark (LSR-Benchmark), a novel dataset targeting real-life scenario reasoning, aiming to close the gap in artificial neural networks' ability to reason in everyday contexts. In contrast to domain knowledge reasoning datasets, LSR-Benchmark comprises free-text formatted questions with rich information on real-life scenarios, human behaviors, and character roles. The dataset consists of 2,162 questions collected from open-source online sources and is manually annotated to improve its quality. Experiments are conducted using state-of-the-art language models, such as gpt3.5-turbo and instruction fine-tuned llama models, to test the performance in LSR-Benchmark. The results reveal that humans outperform these models significantly, indicating a persisting challenge for machine learning models in comprehending daily human life.

Xiezhi: An Ever-Updating Benchmark for Holistic Domain Knowledge Evaluation

New Natural Langauge Process~(NLP) benchmarks are urgently needed to align with the rapid development of large language models (LLMs). We present Xiezhi, the most comprehensive evaluation suite designed to assess holistic domain knowledge. Xiezhi comprises multiple-choice questions across 516 diverse disciplines ranging from 13 different subjects with 249,587 questions and accompanied by Xiezhi-Specialty and Xiezhi-Interdiscipline, both with 15k questions. We conduct evaluation of the 47 cutting-edge LLMs on Xiezhi. Results indicate that LLMs exceed average performance of humans in science, engineering, agronomy, medicine, and art, but fall short in economics, jurisprudence, pedagogy, literature, history, and management. We anticipate Xiezhi will help analyze important strengths and shortcomings of LLMs, and the benchmark is released in~\url{https://github.com/MikeGu721/XiezhiBenchmark}.

Domain Mastery Benchmark: An Ever-Updating Benchmark for Evaluating Holistic Domain Knowledge of Large Language Model--A Preliminary Release

Domain knowledge refers to the in-depth understanding, expertise, and familiarity with a specific subject, industry, field, or area of special interest. The existing benchmarks are all lack of an overall design for domain knowledge evaluation. Holding the belief that the real ability of domain language understanding can only be fairly evaluated by an comprehensive and in-depth benchmark, we introduces the Domma, a Domain Mastery Benchmark. DomMa targets at testing Large Language Models (LLMs) on their domain knowledge understanding, it features extensive domain coverage, large data volume, and a continually updated data set based on Chinese 112 first-level subject classifications. DomMa consist of 100,000 questions in both Chinese and English sourced from graduate entrance examinations and undergraduate exams in Chinese college. We have also propose designs to make benchmark and evaluation process more suitable to LLMs.