laya-finance-matched-sft-200k

Paper: S²-MAS. The fixed-budget adaptation results and matched downstream methods define the comparison reported here.

Project background

S²-MAS is a society-scale multi-agent simulation framework for studying how household decisions affect consumption, cash reserves and payments. The paper connects a constructed resource of 500 million household-persona records, learned decision policies, reusable financial scenarios and a state-transition engine. Its purchase-funding application evaluates 12 million distinct synthetic household agents under paired conditions. Full-data Fin-Laya supplies the policy for that application; this repository supplies a fixed-budget adaptation checkpoint for the common decision-model comparison. Decision learning uses real-survey data. The GENESE/MAS engine implementation remains private.

The benchmark groups twelve systems into six decision-model routes: encoder option scoring; language models with learned decision heads; next-token decision SFT; dynamic-schema classification; multimodal encoder decisions; and provider decision interfaces. All receive text inputs in the common benchmark. The 17 model-repository entries comprise 10 fixed-budget trained variants—five native Laya and five Qwen variants—plus six checkpoint/runtime entries and one compatibility alias.

This model's role

This checkpoint follows the encoder option-scoring route. It adapts the native Laya encoder and its original decision head on the 200k training set. It is also the common parent for the matched RLCD200 and CE-continuation comparisons.

Training design and checkpoint

Method: Supervised fine-tuning (SFT).
Distinct downstream training rows: 200,000 from the fixed decision-QA training partition.
Initialization: convaiinnovations/laya.
This stage terminal row exposures: 200,000; selected checkpoint stage exposures: 200,000.
Selected checkpoint cumulative row exposures: 200,000; selected cumulative distinct rows: 200,000.
A fixed 20,000-example development set selects this checkpoint.

Performance and matched comparisons

Language: English. Inference precision: Native PyTorch BF16.

The common held-out test contains 240,458 QA records, 119 question versions and 8,275 respondent/household groups. The checkpoint is selected using development data before this test.

Checkpoint / reference	Accuracy (%)	Question-macro F1 (%)	Invalid decisions
This checkpoint	73.8753	35.9949	0
Original native Laya, matched BF16 reference	60.1972	29.5995	0

Accuracy weights every test record equally; invalid decisions count as incorrect. Question-macro F1 averages over every defined candidate class within each question version and then weights the 119 versions equally.

Relative to the original native reference, the paper reports +13.6781 accuracy points (95% CI 13.4662–13.8857) and +6.3953 question-macro F1 points (5.9603–6.9003).

Package and inference

Complete fine-tuned encoder and original native decision head.

Initialized from the official task-trained Laya checkpoint, followed by the matched downstream training described above.

Reported common-test inference: native BF16.
Underlying encoder: answerdotai/ModernBERT-large; the original two-layer Laya decision head is retained.
Inference scores the full state, question and complete candidate menu.

Use the native Laya inference interface with the supplied encoder configuration, tokenizer and decision head.

License and access

The project model contribution is offered under the S²-MAS Controlled Model Research License v1.2: noncommercial scientific use, no redistribution, and no unauthorized reverse engineering, extraction or distillation. Training and evaluation records are not distributed. Source-data access requires the provider's authorization. Permission requests: yli9919@hku.hk.

Project repository: s2-mas.