Search Results for author:
Found 9 papers, 6 papers with code
Contrastive Reinforcement Learning of Symbolic Reasoning Domains
Our results suggest new directions for reinforcement learning in symbolic domains, as well as applications to mathematics education.
Open-domain clarification question generation without question examples
An overarching goal of natural language processing is to enable machines to communicate seamlessly with humans.
Synchromesh: Reliable code generation from pre-trained language models
Then, Synchromesh feeds the examples to a pre-trained language model and samples programs using Constrained Semantic Decoding (CSD): a general framework for constraining the output to a set of valid programs in the target language.
LEMMA: Bootstrapping High-Level Mathematical Reasoning with Learned Symbolic Abstractions
Humans tame the complexity of mathematical reasoning by developing hierarchies of abstractions.
Peano: Learning Formal Mathematical Reasoning
We explore this idea in a case study on 5 sections of beginning algebra on the Khan Academy platform.
Parsel: Algorithmic Reasoning with Language Models by Composing Decompositions
Despite recent success in large language model (LLM) reasoning, LLMs struggle with hierarchical multi-step reasoning tasks like generating complex programs.
Ranked #7 on
Code Generation
on HumanEval
Solving Math Word Problems by Combining Language Models With Symbolic Solvers
Automatically generating high-quality step-by-step solutions to math word problems has many applications in education.
Certified Deductive Reasoning with Language Models
In experiments on PrOntoQA, ProofWriter and Syllogism Validity datasets, \textsc{LogicGuide} significantly improves the performance of GPT-3, GPT-3. 5 Turbo and LLaMA (accuracy gains up to 35\%), while drastically reducing \emph{content effects} -- the interference between unwanted prior assumptions and reasoning, which humans and language models suffer from.
Hypothesis Search: Inductive Reasoning with Language Models
Because of the prohibitive cost of generation with state-of-the-art LLMs, we consider a middle step to filter the set of hypotheses that will be implemented into programs: we either ask the LLM to summarize into a smaller set of hypotheses, or ask human annotators to select a subset of the hypotheses.
