AI Paper Atlas

Research map / Large language models / Agents, code and math

From Hypothesis to Premises: LLM-based Backward Logical Reasoning with Selective Symbolic Translation

Qingchuan Li, Mingyue Cheng, Zirui Liu, Daoyu Wang, Yuting Zeng, Tongxuan Liu

AAAI 2026 · Large language models › Agents, code and math · approach: Mathematical and structural reasoning benchmarks · 4 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use backward hypothesis-driven reasoning instead of forward CoT reasoning (ablation)HBLR · ProofWriteraccuracy: 81.24 -> 89.41Improved
use backward hypothesis-driven reasoning instead of forward CoT reasoning (ablation)HBLR · AR-LSATaccuracy: 38.86 -> 44.07Improved
use selective translation instead of all-FOL translation (ablation)HBLR · ProntoQAaccuracy: -1.72% vs All-FOLMixed
use selective translation instead of all-NL translation (ablation)HBLR · five benchmarksaccuracy: +7.2% averageImproved
use selective translation instead of all-FOL translation (ablation)HBLR · five benchmarksaccuracy: +4.7% averageImproved
use selective translation instead of all-FOL translation (ablation)HBLR · ProofWriteraccuracy: +12.83%Improved
HBLR vs SymbCoT on GPT-4HBLR · FOLIOaccuracy: 78.19 -> 84.22Improved
HBLR vs Logic-LM on GPT-3.5-TurboHBLR · LogicalDeductionaccuracy: 69.06 -> 49.77Hurt
HBLR vs CoT on DeepSeek-R1HBLR · AR-LSATaccuracy: 76.91 -> 86.47Improved
HBLR vs CoT on DeepSeek-R1HBLR · ProntoQAaccuracy: 99.57 -> 99.72No clear effect
backward reasoning raises effective reasoning rate vs SymbCoTHBLR · ProntoQAeffective reasoning rate: 73.65 -> 95.58Improved
upgrade DeepSeek-V3 to DeepSeek-R1 under HBLRHBLR · average of benchmarksaccuracy: 82.31 -> 94.34Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy