PsyMemBehav: Psychology-Informed and Memory-Grounded Two-Level Identity-to-Action Reasoning for

We introduce PsyMemBehav, which maps psychology-informed profiles and retrieved memories to target-first platform behavior through two-level identity-to-action reasoning. It comes from a paper posted to arXiv on 15 October 2026. Large language models (LLMs) support social simulation, but faithfully reproducing real users requires grounding both structured platform decisions and language in sparse, noisy behavioral evidence. To support this approach with comprehensive behavioral traces, we construct a dataset PsyMemBehav-bench from 5.57M observed behaviors of 5,862 real users (approximately 950 per user) across Reddit, Bluesky, and Twitter. Each recorded engagement is modeled as target selection followed by Comment, Like, or Repost and optional response generation. PsyMemBehav combines 41-dimensional user descriptors, top-5 BM25 memories, reverse-CoT supervision, and rubric-guided DPO over complete reasoning-action trajectories. Experiments show that PsyMemBehav 27B ranks first in 11 of 12 reported individual-level metrics, improving average response quality over GPT-5.4 from 54.28 to 59.68 (9.95% relative), and obtains the lowest group-level MAE (0.081) and MSE (0.009) among simulated systems. These results support evidence-grounded user modeling for observed-thread continuation; the benchmark remains conditional on recorded engagement and does not model no-action decisions.

Why it counts

We introduce PsyMemBehav, which maps psychology-informed profiles and retrieved memories to target-first platform behavior through two-level identity-to-action reasoning. Experiments show that PsyMemBehav 27B ranks first in 11 of 12 reported individual-level metrics, improving average response quality over GPT-5.4 from 54.28 to 59.68 (9.95% relative), and obtains the lowest group-level MAE (0.081) and MSE (0.009) among simulated systems.

Sources

Primary source: primary source

What is not known

This brief does not claim independent replication. Claims that appear only on X and not in the primary source stay unknown.