AI Paper Atlas

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

Distilling Neuro-Symbolic Programs into 3D Multi-modal LLMs

Wentao Mo, Yang Liu

ICML 2026 · Large language models › Agents, code and math · approach: Inductive and abstract reasoning evaluation · 0 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
add CoT-RL stage (Stage 3) instead of direct-answer SFT (ablation)APEIRIA · ScanRefer Acc@0.25accuracy: 51.5 -> 58.4Improved
add CoT-SFT symbolic reasoning injection (Stage 2) before RL (ablation)APEIRIA · Multi3DRefer F1@0.25F1: 36.7 -> 59.2Improved
add format reward in RL (ablation)APEIRIA · ScanRefer Acc@0.25accuracy: 55.7 -> 58.4Improved
use soft grounding reward instead of IoU reward (ablation)APEIRIA · ScanRefer Acc@0.25accuracy: 57.7 -> 58.4Improved
generate thinking trace at inference instead of direct answer (ablation)APEIRIA · ScanRefer Acc@0.25accuracy: 56.8 -> 58.4Improved
CoT-RL on short reasoning chains (<=4 steps) (ablation)APEIRIA · ScanRefer Acc@0.5accuracy: 47.2 -> 45.4Hurt
CoT-RL on medium chains (5 steps) (ablation)APEIRIA · ScanRefer Acc@0.5accuracy: 50.7 -> 52.2Improved
CoT-RL on long chains (>=6 steps) (ablation)APEIRIA · ScanRefer Acc@0.5accuracy: 45.2 -> 47.9Improved
inject plans from Claude 4.5 Opus instead of self-generated plans (ablation)APEIRIA · ScanRefer Acc@0.25accuracy: 58.4 -> 58.6No clear effect
replace perception with SegDINO3D outputs (ablation)APEIRIA · ScanRefer Acc@0.25accuracy: 58.4 -> 60.4Improved
replace perception with Mask3D outputs (ablation)APEIRIA · ScanRefer Acc@0.25 / Multi3DRefer F1@0.25accuracy / F1: 58.4 -> 58.5; 59.2 -> 58.3Mixed
train on Sr3D only and test zero-shot on Nr3DAPEIRIA vs supervised NS3D · Nr3Daccuracy: 33.9 -> 36.5Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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