An official website of the United States government
Large language models (LLMs) have demonstrated an impressive ability to perform arithmetic and symbolic reasoning tasks, when provided with a few examples at test time ("few-shot prompting"). Much of this success can be attributed to prompting methods such as "chain-of-thought", which employ LLMs for both understanding the problem description by decomposing it into steps, as well as solving each step of the problem. While LLMs seem to be adept at this sort of step-by-step decomposition, LLMs often make logical and arithmetic mistakes in the solution part, even when the problem is decomposed correctly. In this paper, we present Program-Aided Language models (PAL): a novel approach that uses the LLM to read natural language problems and generate programs as the intermediate reasoning steps, but offloads the solution step to a runtime such as a Python interpreter. With PAL, decomposing the natural language problem into runnable steps remains the only learning task for the LLM, while solving is delegated to the interpreter. We demonstrate this synergy between a neural LLM and a symbolic interpreter across 13 mathematical, symbolic, and algorithmic reasoning tasks from BIG-Bench Hard and others. In all these natural language reasoning tasks, generating code using an LLM and reasoning using a Python interpreter leads to more accurate results than much larger models. For example, PAL using Codex achieves state-of-the-art few-shot accuracy on GSM8K, surpassing PaLM which uses chain-of-thought by absolute 15% top-1.
more »
« less
Improving Coherence and Consistency in Neural Sequence Models with Dual-System, Neuro-Symbolic Reasoning
Human reasoning can often be understood as an interplay between two systems: the intuitive and associative (“System 1”) and the deliberative and logical (“System 2”). Neural sequence models—which have been increasingly successful at performing complex, structured tasks—exhibit the advantages and failure modes of System 1: they are fast and learn patterns from data, but are often inconsistent and incoherent. In this work, we seek a lightweight, training-free means of improving existing System 1-like sequence models by adding System 2-inspired logical reasoning. We explore several variations on this theme in which candidate generations from a neural sequence model are examined for logical consistency by a symbolic reasoning module, which can either accept or reject the generations. Our approach uses neural inference to mediate between the neural System 1 and the logical System 2. Results in robust story generation and grounded instruction-following show that this approach can increase the coherence and accuracy of neurally-based generations.
more »
« less
- Award ID(s):
- 1911790
- PAR ID:
- 10279463
- Date Published:
- Journal Name:
- ArXivorg
- ISSN:
- 2331-8422
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abstract In this paper, we evaluate the capability of transformer-based language models in making inferences over uncertain text that includes uncertain rules of reasoning. We cover both Pre-trained Language Models (PLMs) and generative Large Language Models (LLMs). Our evaluation results show that both generations of language models struggle with reasoning over uncertain text. We propose a novel end-to-end fine-tuning approach, Probabilistic Constraint Training (PCT), that utilizes probabilistic logical rules as constraints in the fine-tuning phase without relying on these rules in the inference stage. To assess the effectiveness of PCT, we utilize the related corpora and, additionally, create a new and more challenging benchmark that, unlike the previous ones, uses instance-specific rules. Our study demonstrates that PCT improves the transformer-based language model’s intrinsic reasoning and makes their probabilistic logical reasoning process more explicit and explainable. Furthermore, PCT equips these models to effectively handle novel situations, including higher reasoning depth, new domains, and complex probabilistic structures.more » « less
-
Neural models, including large language models (LLMs), achieve superior performance on logical reasoning tasks such as question answering. To elicit reasoning capabilities from LLMs, recent works propose using the chain-of-thought (CoT) mechanism to generate both the reasoning chain and the answer, which enhances the model’s capabilities in conducting reasoning. However, due to LLM’s uninterpretable nature and the extreme flexibility of free-form explanations, several challenges remain: such as struggling with inaccurate reasoning, hallucinations, and not aligning with human preferences. In this talk, we will focus on (1) our design of leveraging structured information (that is grounded to the context), for the explainable complex question answering and reasoning; (2) our multi-module interpretable framework for inductive reasoning, which conducts step-wise faithful reasoning with iterative feedback.more » « less
-
Deep learning and symbolic reasoning are complementary techniques for an intelligent system. However, principled combinations of these techniques are typically limited in scalability, rendering them ill-suited for real-world applications. We propose Scallop, a system that builds upon probabilistic deductive databases, to bridge this gap. The key insight underlying Scallop is a provenance framework that introduces a tunable parameter to specify the level of reasoning granularity. Scallop thereby i) generalizes exact probabilistic reasoning, ii) asymptotically reduces computational cost, and iii) provides relative accuracy guarantees. On synthetic tasks involving mathematical and logical reasoning, Scallop scales significantly better without sacrificing accuracy compared to DeepProbLog, a principled neural logic programming approach. Scallop also scales to a newly created real-world Visual Question Answering (VQA) benchmark that requires multi-hop reasoning, achieving 84.22% accuracy and outperforming two VQA-tailored models based on Neural Module Networks and transformers by 12.42% and 21.66% respectively.more » « less
-
Understanding an online argumentative discussion is essential for understanding users' opinions on a topic and their underlying reasoning. A key challenge in determining completeness and persuasiveness of argumentative discussions is to assess how arguments under a topic are connected in a logical and coherent manner. Online argumentative discussions, in contrast to essays or face-to-face communication, challenge techniques for judging argument relevance because online discussions involve multiple participants and often exhibit incoherence in reasoning and inconsistencies in writing style. We define relevance as the logical and topical connections between small texts representing argument fragments in online discussions. We provide a corpus comprising pairs of sentences, labeled with argumentative relevance between the sentences in each pair. We propose a computational approach relying on content reduction and a Siamese neural network architecture for modeling argumentative connections and determining argumentative relevance between texts. Experimental results indicate that our approach is effective in measuring relevance between arguments, and outperforms strong and well-adopted baselines.Further analysis demonstrates the benefit of using our argumentative relevance encoding on a downstream task, predicting how impactful an online comment is to certain topic, comparing to encoding that does not consider logical connection.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- The DOI is not currently available.
