AI Paper Atlas

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

OmniSimulator: Aligning Small Language Models for Authentic Heterogeneous Behavior Modeling

Jiawei Chen, Ruoxi Xu, Boxi Cao, Zhuoqun Li, Ruotong Pan, Zhang yunfei, Zhirui Yang, Jingyan Chen, Tingting Gao, Han Li, Yaojie Lu, Xianpei Han and 3 more

NeurIPS 2026 · Large language models › Agents, code and math · approach: Mathematical and structural reasoning benchmarks · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes OmniSimulator, a method for aligning small language models to act as user simulators that model heterogeneous user behavior more authentically than standard chain-of-thought.

Problem

LLM-based user simulation for interactive systems is limited by the model's reasoning paradigm; standard chain-of-thought tends to degenerate into shallow patterns, capping simulation fidelity.

Approach

Aligns small language models for heterogeneous user behavior modeling, moving beyond standard chain-of-thought reasoning to improve authenticity of simulated users; specific mechanism not stated in the abstract.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
enable native thinking modeQwen3-8B zero-shot · OmniBehavioraverage score: 30.82 -> 30.77No clear effect
enable native thinking modeQwen3-235B zero-shot · OmniBehavioraverage score: 33.15 -> 33.79No clear effect
add SFT on free-form teacher CoTQwen3-8B · OmniBehavioraverage score: 30.82 -> 34.34Improved
use structured five-dimension CoT with rejection-sampled distillation instead of free CoT (ablation)Qwen3-8B · OmniBehavioraverage score: 34.34 -> 41.40Improved
add rejection sampling to teacher traces (ablation)OmniSimulator Qwen3-8B · OmniBehavioraverage score: 37.44 -> 41.40Improved
train on answers only instead of reasoning traces (ablation)Qwen3-8B · OmniBehavioraverage score: 30.82 -> 31.62No clear effect
use generic synthetic rationales instead of the structured schema (ablation)Qwen3-8B · OmniBehavioraverage score: 41.40 -> 37.85Hurt
include Historical Behavioral State dimension (ablation)OmniSimulator Qwen3-1.7B · OmniBehavior binaryF1: 19.65 -> 28.80Improved
include User Profile Grounding dimension (ablation)OmniSimulator Qwen3-1.7B · OmniBehavior textualLLM-judge score: 31.99 -> 40.98Improved
include Decision Mode Identification dimension (ablation)OmniSimulator Qwen3-1.7B · OmniBehavior binaryF1: 23.42 -> 28.80Improved
include Content Relevance Alignment dimension (ablation)OmniSimulator Qwen3-1.7B · OmniBehavior continuous100 x (1-NMAE): 60.21 -> 62.12Improved
include all five schema dimensions versus none (ablation)OmniSimulator Qwen3-1.7B · OmniBehavior binaryF1: 21.78 -> 28.80Improved

Experimental setup

Baselines: standard chain-of-thought

Builds on: chain-of-thought

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