S$^2$-MAS: A Society-Scale Multi-Agent Simulation Framework with 500 Million Household Persona Agents

Yang Li; Yike Wang; Di Wan; Chujun He; Zongsi Xu; Yitao Yu; Quanfeng Lu; Xiyu Chen

cs.MAOct 20, 2026

S $^2$ -MAS: A Society-Scale Multi-Agent Simulation Framework with 500 Million Household Persona Agents

Authors: Yang Li, Yike Wang, Di Wan, Chujun He, Zongsi Xu, Yitao Yu, Quanfeng Lu, Xiyu Chen,

Organizations: The University of Hong Kong · Faculty of Business and Economics, The University of Hong Kong · HKU Business School Shenzhen Research Institutes, The University of Hong Kong · Nanyang Technological University ·

Abstract

Simulating economic responses at society scale requires agents whose decisions reflect household circumstances. We introduce S $^2$ -MAS, a society-scale multi-agent simulation framework built on 500 million synthetic household personas grounded in real-survey data. Its Genese engine separates decisions from state updates, while reusable financial scenarios and application ledgers track consumption, cash, debt and payments. Trained with SFT + RLCD on 2.93 million examples, sharp-421m-mlx achieves 74.55% accuracy and 38.37% question-macro F1 on one-shot consumer financial-choice tasks derived from held-out survey responses. Its accuracy exceeds GLM-5.3-Flash, Jev 1.13, DeepSeek-V4.1-Flash, GPT-6-Luna and GPT-6-Astra in the common benchmark (full test for the first four; matched 500-group sample for Astra). We evaluate eight financial task families and simulate 12 million distinct household agents in paired purchase-funding conditions. Jointly raising purchase goals and approved credit limits increases consumption, with responses differing across age and income groups. Education, employment and household-size differences persist after age-income standardization and mainly reflect transitions from deferral to purchase. Asset and debt groups differ in cash drawdown, repayment and closing debt. S $^2$ -MAS links heterogeneous household decisions to evolving consumption, cash and repayment paths, supporting economic experiments across household groups and population scales.

Figures & tables

Task / changed condition	Offered actions	Financial measures	Evidence reported here
T1: income and consumption	Adjust discretionary spending	Consumption, cash, unmet needs	Paired static sharp-421m/Laya execution (Appendix S11)
T2: medical expense funding	Use savings, borrow or combine both	Cash, debt, unmet needs	paired sharp-421m/Laya execution (Appendix S11)
T3: purchase-funding need	Defer purchase or finance a partial/full amount	Consumption, cash, payments, debt	Paired bilingual execution and 12M application (§6.2, Tables S32–S33)
T4: loan product and term	Choose a short/long contract or defer purchase	Payments, cash, arrears	paired sharp-421m/Laya execution (Appendix S11)
T5: liquid saving and risky assets	Allocate between liquid and risky holdings	Cash, assets, unfunded needs	paired sharp-421m/Laya execution (Appendix S11)
T6: multiple-debt payments	Allocate a payment budget across debts	Debt balances, arrears, default	paired sharp-421m/Laya execution (Appendix S11)

Table 2

Figure 3

Figure 4

Figure 6

Figure 7

Model / adaptation	Unique training QA	Selected cumulative exposures	Accuracy (%)	Question-macro F1 (%)
Original native Laya	0	0	60.1972	29.5995
sharp-4b-head-50k	50,000	100,000	67.3369	31.0166
sharp-4b-lora-50k	50,000	50,000	73.7289	36.6225
sharp-4b-head-200k	200,000	200,000	67.2841	30.8995
sharp-4b	200,000	200,000	74.1435	36.7343
sharp-9b	200,000	200,000	74.0919	36.6622

Table 8

Model / information condition	QA	Weighted accuracy (%)	Question-macro F1 (%)
Base Laya, full profile	20000	60.14	29.26
sharp-421m, full profile	20000	74.40	37.77
sharp-421m, no profile	20000	71.19	31.53
sharp-421m, wrong same-question profile	20000	69.56	34.17
Training-question majority	20000	71.71	32.09

Table 9

Model	Model size	Backend	QA / valid	Accuracy (%)	Question-macro F1 (%)
sharp-421m-mlx	421M	MLX CUDA FP16	240,458 / 240,458	74.5544	38.3746
Question-majority baseline	—	CPU; training labels	240,458 / 240,458	71.7710	32.1688
GLM-5.3-Flash	320B	API; provider precision	240,458 / 240,360	68.8756	37.9991
Jev 1.13	Not disclosed	API; provider precision	240,458 / 240,458	65.2584	37.4793
Kev 4B	4B	CUDA BF16	240,458 / 240,458	64.9120	36.0435
DeepSeek-V4.1-Flash	552B + 196B	Provider API	240,458 / 240,400	63.7546	37.9400

Table 10

Tested component and contrast	Observed result	Contribution to the framework
Respondent information: full versus empty profile, same sharp-421m policy and questions; 20,000 QA	Full profiles add 6.2475 points of question-macro F1 (95% interval 5.4958–6.7267)	Agent attributes contribute predictive information beyond the question alone
Compact domain adaptation: original native Laya versus SFT on 200k unique QA; same 421M architecture, BF16	Accuracy rises from 60.1972% to 73.8753%	Survey supervision supplies most of the native policy’s downstream gain
Continuation objective: same SFT200 parent and added 200k-example pass budget; development-selected RLCD and CE checkpoints	RLCD raises accuracy from SFT’s 73.8753% to 73.9268%. Against CE continuation, RLCD’s F1 is 0.1229 points lower (95% interval -0.2073 to -0.0330)	The continuation objective changes the metric balance after SFT; CE has higher F1 in this comparison
Representation adaptation: sharp-4b-head-200k versus sharp-4b on the same 200k QA	Accuracy is 67.2841% with head-only training and 74.1435% with LoRA	Updating the representation adds value beyond the frozen-backbone route
Backbone substitution: sharp-4b and sharp-9b, same 200k-example one-pass LoRA protocol; different upstream histories	Accuracy is 74.1435% and 74.0919%; 9B minus 4B is -0.0516 points (-0.0981 to -0.0071)	These two adaptation routes reach similar accuracy on the common test
External policy substitution: full-data sharp-421m versus complete decision-model systems on common records and alternatives	sharp-421m exceeds Jev by 9.2960 accuracy points and 0.8953 F1 points. Clef has the highest F1: 38.5193%, versus sharp-421m’s 38.3746%; their paired F1 difference includes zero	A common interface exposes accuracy/F1 tradeoffs across decision systems with different readouts

Table 11

Figure 12

Measurement	Configuration / scope	Genese result	Other results
1,000 agents, immediate backend	2 $\times$ Xeon Platinum 8268; 96 threads; 20 interleaved blocks	Default 2.788 s; partitions 2.743 s	AgentSociety 5.818; OneSim 6.223; OASIS 7.386; MiroFish 11.255; MatrAIx 1,062.153 s
100 agents, immediate backend	2 $\times$ Xeon Platinum 8268; 96 threads; 20 interleaved blocks	Default 0.362 s; partitions 0.360 s	AgentSociety 1.063; OneSim 0.521; OASIS 0.931; MiroFish 1.163; MatrAIx 96.529 s
100 agents, sharp-421m fl4	3 T4 model replicas; 3 blocks	0.517 s	OneSim 0.683; OASIS 1.140; MiroFish 1.403 s
100 agents, sharp-421m fl5	Separate adapter matrix; 3 blocks	0.509 s	MatrAIx uses this separate reference (Supplementary Figure S5d)
1,000 agents, sharp-421m	1 T4 model replica; 2 blocks	6.17 s	OASIS 7.45; OneSim 7.70 s
1,000 agents, Jev	2 blocks	5.97 s	OASIS 8.21 s; OneSim 0/2 completed

Table 13

Figure 14

Figure 15

Agent-Based Modeling

Agent-based modeling simulates how interacting, heterogeneous entities and their environments produce patterns in social, economic, biological, and engineered systems.

See all 160 papers

MALLES: A Multi-agent LLMs-based Economic Sandbox with Consumer Preference Alignment
Mar 18, 2026Yusen Wu, Yiran Liu, Xiaotie DengMulti-Agent LLM Systems Human Behavior Simulation
Poor Man's Agentic Modeling: Simulating Large LLM-Agent Societies on a Laptop
Jul 19, 2026Igor ItkinMulti-Agent LLM Systems Surrogate Modeling
Independent Researcher, Tel Aviv, Israel
FraudSociety: A Generative Agent-Based Society for the Financial-Fraud Economy
Oct 24, 2026Keyang Chen, Mingxuan Jiang, Zeping Li +10Financial Fraud Detection Agent-Based Modeling
Fudan University, Shanghai, China · University of Illinois Urbana-Champaign, Urbana, IL, USA

Mar 18, 2026J/K move · Enter open · S save

MALLES: A Multi-agent LLMs-based Economic Sandbox with Consumer Preference Alignment

Yusen Wu, Yiran Liu, Xiaotie Deng

Multi-Agent LLM Systems Human Behavior Simulation Large Language Model-Based Social Simulation Preference Learning

Open paper AskPDF arXiv

In the real economy, modern decision-making is fundamentally challenged by high-dimensional, multimodal environments, which are further complicated by agent heterogeneity and combinatorial data sparsity. This paper introduces a Multi-Agent Large Language Model-based Economic Sandbox (MALLES), leveraging the inherent generalization capabilities of large-sacle models to establish a unified simulation framework applicable to cross-domain and cross-category scenarios. Central to our approach is a preference learning paradigm in which LLMs are economically aligned via post-training on extensive, heterogeneous transaction records across diverse product categories. This methodology enables the models to internalize and transfer latent consumer preference patterns, thereby mitigating the data sparsity issues prevalent in individual categories. To enhance simulation stability, we implement a mean-field mechanism designed to model the dynamic interactions between the product environment and customer populations, effectively stabilizing sampling processes within high-dimensional decision spaces. Furthermore, we propose a multi-agent discussion framework wherein specialized agents collaboratively process extensive product information. This architecture distributes cognitive load to alleviate single-agent attention bottlenecks and captures critical decision factors through structured dialogue. Experiments demonstrate that our framework achieves significant improvements in product selection accuracy, purchase quantity prediction, and simulation stability compared to existing economic and financial LLM simulation baselines. Our results substantiate the potential of large language models as a foundational pillar for high-fidelity, scalable decision simulation and latter analysis in the real economy based on foundational database.

Abstract

Agent-Based Modeling

Explore similar work

MALLES: A Multi-agent LLMs-based Economic Sandbox with Consumer Preference Alignment

Poor Man's Agentic Modeling: Simulating Large LLM-Agent Societies on a Laptop

FraudSociety: A Generative Agent-Based Society for the Financial-Fraud Economy