---
type: "raw_signal"
id: "kaggle-and-rehearsals-launched-human-behavior-bench-scoring-4d93b6"
slug: "kaggle-and-rehearsals-launched-human-behavior-bench-scoring-4d93b6"
title: "Kaggle and Rehearsals launched Human Behavior Bench, scoring LLMs on predicting real human survey distributions."
description: "Rehearsals, Google DeepMind and Kaggle have introduced Human Behavior Bench, a benchmark for predicting real human survey-response distributions for defined audiences."
retrieval_nugget: "Rehearsals, Google DeepMind and Kaggle have introduced Human Behavior Bench, a benchmark for predicting real human survey-response distributions for defined audiences."
observed_at: "2026-10-21"
record_date: "2026-10-21"
date_kind: "observed_at"
why_it_matters: "In the published results, six language models score between 76.2 and 80.6 distribution-agreement points. Rehearsals reports 94.1 for its Engram 1.0 system. The source says base models ran through a Kaggle harness, while Engram was assessed using Rehearsals' in-house harness with the same case and metric setup. The authors also report that LLMs select the top answer correctly 41% of the time on pro"
novelty: "notable"
verification_level: "source-linked"
signal_type: "release"
source_platform: "runrehearsals.com"
topics: ["google"]
entities: ["Rehearsals","Google DeepMind","Kaggle","Human Behavior Bench","Engram"]
related_patterns: []
novelty_label: "new_fact"
what_changed: "Rehearsals, Google DeepMind and Kaggle introduce a distinct benchmark for predicting human survey-response distributions, reporting 94.1% distribution agreement."
prior_refs: ["memory:7df0c1bd-43e6-859f-a008-8709f5030dc9","memory:e752bf6c-a186-8692-af09-0a10f86e07b2"]
source_url: "https://www.runrehearsals.com/human-behavior-bench"
source_urls: ["https://www.runrehearsals.com/human-behavior-bench"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "signal:kaggle-and-rehearsals-launched-human-behavior-bench-scoring-4d93b6"
status: "published"
visuals: []
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/signals/kaggle-and-rehearsals-launched-human-behavior-bench-scoring-4d93b6/","markdown":"https://newruntime.com/signals/kaggle-and-rehearsals-launched-human-behavior-bench-scoring-4d93b6.md","json":"https://newruntime.com/signals/kaggle-and-rehearsals-launched-human-behavior-bench-scoring-4d93b6.json"}
---

# Kaggle and Rehearsals launched Human Behavior Bench, scoring LLMs on predicting real human survey distributions.

## Retrieval answer

Rehearsals, Google DeepMind and Kaggle have introduced Human Behavior Bench, a benchmark for predicting real human survey-response distributions for defined audiences.

Rehearsals, Google DeepMind and Kaggle have introduced Human Behavior Bench, a benchmark for predicting real human survey-response distributions for defined audiences. Version 1.0 reports 1,199 evaluation tasks based on a panel of 1,244 U.S. adults surveyed from June to September 2026; 239 tasks are public and 960 are held out. Each task asks a system to predict a distribution over fixed answer options, scored against ground truth using total variation distance.

In the published results, six language models score between 76.2 and 80.6 distribution-agreement points. Rehearsals reports 94.1 for its Engram 1.0 system. The source says base models ran through a Kaggle harness, while Engram was assessed using Rehearsals' in-house harness with the same case and metric setup. The authors also report that LLMs select the top answer correctly 41% of the time on product questions, versus 74% on brand-perception questions. These are results reported by the benchmark's organizers, not an independent replication.

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- topic hub: [Google](https://newruntime.com/topics/google.md): Explore the google topic hub.
- collection: [Raw Signals](https://newruntime.com/signals.json): Open every source-linked raw signal.
- collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)