LLMs reasoning // knowledge representation
For a man with a hammer (LLM), everything starts to look like nails (texts)
Natural language is the most universal form of knowledge representation, use it wisely. LLMs are unable to generate or synthesize new knowledge, they just can repeat already pre-trained sequences/patterns (not always correctly — hallucinate) as all ML does. Don’t expect the magic!
The prediction of protein-protein interactions (PPIs) is crucial for understanding biological functions and diseases. Previous machine learning approaches to PPI prediction mainly focus on direct physical interactions, ignoring the broader context of nonphysical connections through intermediate proteins, thus limiting their effectiveness. The emergence of Large Language Models (LLMs) provides a new opportunity for addressing this complex biological challenge. By transforming structured data into natural language prompts, we can map the relationships between proteins into texts. This approach allows LLMs to identify indirect connections between proteins, tracing the path from upstream to downstream. Therefore, we propose a novel framework ProLLM that employs an LLM tailored for PPI for the first time. Specifically, we propose Protein Chain of Thought (ProCoT), which replicates the biological mechanism of signaling pathways as natural language prompts. ProCoT considers a signaling pathway as a protein reasoning process, which starts from upstream proteins and passes through several intermediate proteins to transmit biological signals to downstream proteins. Thus, we can use ProCoT to predict the interaction between upstream proteins and downstream proteins. The training of ProLLM employs the ProCoT format, which enhances the model’s understanding of complex biological problems. In addition to ProCoT, this paper also contributes to the exploration of embedding replacement of protein sites in natural language prompts, and instruction fine-tuning in protein knowledge datasets. We demonstrate the efficacy of ProLLM through rigorous validation against benchmark datasets, showing significant improvement over existing methods in terms of prediction accuracy and generalizability. Our results highlight the potential of LLMs to transform the field of PPI, serving as a robust potential tool for various categories of biological and medical research. The code is available at: https://github.com/MingyuJ666/ProLLM.
This study investigates whether OpenAI’s ChatGPT-3.5 and ChatGPT-4 can accurately forecast future events using two distinct prompting strategies. To evaluate the accuracy of the predictions, we take advantage of the fact that the training data at the time of our experiments (April and May 2023) stopped at September 2021, and ask about events that happened in 2022. We employed two prompting strategies: direct prediction and what we call future narratives which ask ChatGPT to tell fictional stories set in the future with characters that share events that have happened to them, but after ChatGPT’s training data had been collected. Concentrating on events in 2022, we prompted ChatGPT to engage in storytelling, particularly within economic contexts. After analyzing 100 prompts, we discovered that future narrative prompts significantly enhanced ChatGPT-4’s forecasting accuracy. This was especially evident in its predictions of major Academy Award winners as well as economic trends, the latter inferred from scenarios where the model impersonated public figures like the Federal Reserve Chair, Jerome Powell. As a falsification exercise, we repeated our experiments in May 2024 at which time the models included more recent training data. ChatGPT-4’s accuracy significantly improved when the training window included the events being prompted for, achieving 100% accuracy in many instances. The poorer accuracy for events outside of the training window suggests that in the 2023 prediction experiments, ChatGPT-4 was forming predictions based solely on its training data. Narrative prompting also consistently outperformed direct prompting. These findings indicate that narrative prompts leverage the models’ capacity for hallucinatory narrative construction, facilitating more effective data synthesis and extrapolation than straightforward predictions. Our research reveals new aspects of LLMs’ predictive capabilities and suggests potential future applications in analytical contexts.
Abstract — As the dependence on computer systems expands across various domains, focusing on personal, industrial, and large-scale applications, there arises a compelling need to enhance their reliability to sustain business operations seamlessly and ensure optimal user satisfaction. Achieving enhanced reliability depends on predicting and preemptively addressing these failures presents a formidable challenge. However, the challenge of predicting failures and mitigate them in timely manner is a complex problem. Thus, it becomes important to adopt effective strategies for managing failures and minimizing their impact. System logs generated by these devices serve as valuable repositories of historical trends and past failures. The use of machine learning techniques for failure prediction has become commonplace, enabling the extraction of insights from past data to anticipate future behavior patterns. Recently, large language models have demonstrated remarkable capabilities in tasks including summarization, reasoning, and event prediction. Therefore, in this paper, we endeavor to investigate the potential of large language models in predicting system failures, leveraging insights learned from past failure behavior to inform reasoning and decision-making processes effectively. Our approach involves leveraging data from the Intel® Computing Improvement Program (ICIP) system crash logs to identify significant events and develop CrashEventLLM. This model, built upon a large language model framework, serves as our foundation for crash event prediction. Specifically, our model utilizes historical data to forecast future crash events, informed by expert annotations. Additionally, it goes beyond mere prediction, offering insights into potential causes for each crash event. This work provides the preliminary insights into prompt-based large language models for the log-based event prediction task. Index Terms — log data, log analysis, large language model, crash prediction using LLM, deep learning, machine learning.
Abstract — Large language models for code (i.e., code LLMs) have shown strong code understanding and generation capabilities. To evaluate the capabilities of code LLMs in various aspects, many benchmarks have been proposed (e.g., HumanEval and ClassEval). Code reasoning is one of the most essential abilities of code LLMs, but existing benchmarks for code reasoning are not sufficient. Typically, they focus on predicting the input and output of a program, ignoring the evaluation of the intermediate behavior during program execution, as well as the logical consistency (e.g., the model should not give the correct output if the prediction of execution path is wrong) when performing the reasoning. To address these problems, in this paper, we propose a framework, namely REval, for evaluating code reasoning abilities and consistency of code LLMs with program execution. We utilize existing code benchmarks and adapt them to new benchmarks within our framework. A large-scale empirical study is conducted and most LLMs show unsatisfactory performance on both Runtime Behavior Reasoning (i.e., an average accuracy of 44.4%) and Incremental Consistency Evaluation (i.e., an average IC score of 10.3). Evaluation results of current code LLMs reflect the urgent need for the community to strengthen the code reasoning capability of code LLMs. Our code, data, and REval leaderboard are available at https://r-eval.github.io.
Traditional language model-based theorem proving assumes that by training on a sufficient amount of formal proof data, a model will learn to prove theorems. Our key observation is that a wealth of informal information that is not present in formal proofs can be useful for learning to prove theorems. For instance, humans think through steps of a proof, but this thought process is not visible in the resulting code. We present Lean-STaR, a framework for training language models to produce informal thoughts prior to each step of a proof, thereby boosting the model’s theorem-proving capabilities. Lean-STaR uses retrospective ground-truth tactics to generate synthetic thoughts for training the language model. At inference time, the trained model directly generates the thoughts prior to the prediction of the tactics in each proof step. Building on the self-taught reasoner framework, we then apply expert iteration to further fine-tune the model on the correct proofs it samples and verifies using the Lean solver. Lean-STaR achieves state-of-the-art results on the miniF2F-test benchmark within the Lean theorem proving environment, significantly outperforming base models (43.4% → 46.3%, Pass@64). We also analyze the impact of the augmented thoughts on various aspects of the theorem proving process, providing insights into their effectiveness.
LLMs excel at information-seeking and creative writing tasks. They have significantly improved in math and coding too. But how do they perform in logical reasoning?
To evaluate the logical reasoning abilities of LLMs, we have created a benchmark named ZebraLogic. Each example is a Logic Grid Puzzle, also known as a Zebra Puzzle. In each puzzle, we are given N houses (numbered 1 to N from left to right) and M features for each house. There are N distinct values for each feature, and each house must have a unique value for each feature. Given a list of clues, one should be able to deduce a unique correct assignment of values. The logic grid puzzle is a typical Constraint Satisfaction Problem (CSP) and is often used to test humans’ logical reasoning abilities in exams such as the Law School Admission Test (LSAT).
