Small Foundation Models of Human Cognition and Behaviour

arXiv:2608.052242026-08-04

Small models with only a few hundred million to a billion parameters matched a 70-billion-parameter model at predicting human choices, as long as the test was similar to what they trained on

Researchers fine-tuned 14 language models, ranging from 135 million to 14 billion parameters, on Psych-101, a dataset of 10.7 million trial-level human choices from 160 psychology experiments, to test whether model size actually matters and whether these models use real task information or just statistical shortcuts. Within the same kind of experiments they trained on, size barely mattered, and models with 0.6 to 1 billion parameters matched a reproduced 70-billion-parameter baseline, but on entirely new, unseen experiments larger models clearly generalised better. By systematically stripping parts of the prompt, the team showed the models rely on the actual content of stimuli and feedback shown in each experiment, not just a memorised history of past choices.

METAL MEDIA explanatory visual

Three-stage test of small cognitively fine-tuned models

Evidence statusMeasured results reported

1. Scale sweep14 models from 135M to 14B parameters fine-tuned on Psych-101 (160 experiments, 10.7M trials), compared in-distribution and out-of-distribution
2. Prompt channel ablationTask instructions, stimuli, feedback, and choice history removed one at a time across 27 experiments to see which information actually drives predictions
3. Trial order permutationTrial order shuffled in an independent-trial task (THINGS) and a sequentially dependent task (intertemporal choice) to test whether models respect each task's structure
Conclusion: task-adaptive information useThe models rely on the actual content of stimuli and feedback shown in each experiment, adapted to that task's structure, rather than on memorised history or prompt formatting alone

An explanatory diagram made by METAL MEDIA, not a figure supplied by the paper's authors.

What they did

The team fine-tuned 14 models (135M to 14B parameters) from four model families (Llama, Qwen3, SmolLM, OLMo) on Psych-101, a dataset of 10.7 million trial-level human choices drawn from 160 psychology experiments.
On experiments similar to the training data (in-distribution), model size barely mattered: models with only 0.6 to 1 billion parameters matched their reproduction of the 70-billion-parameter Centaur model, but on entirely new experiments (out-of-distribution) larger models generalised clearly better.
To find out what information the models actually rely on, the researchers split each prompt into four parts, task instructions, the experimental stimuli shown, outcome feedback, and choice history, and removed them one at a time across 27 experiments.
Masking just the specific content of stimuli and feedback, while keeping the same formatting, destroyed 75.7% of the information the models had learned and pushed accuracy below random-guessing level, showing the models are not simply memorising past choice patterns.
Shuffling the order of trials left predictions unchanged on tasks where trials are independent of each other, but changed predictions on tasks where each trial depends on the previous response, showing the models adapt how they use information to match each task's actual structure.

Figure 1: Adapter rank against model size on Psych-101. Mean negative log-likelihood over the 38 of 46 Psych-101 tasks for which 11 publish a domain-specific cognitive model, plotted against parameter count, one panel per adapter rank (r=4–64). Trend lines are fitted only within groups matched on generation and context window, so Olmotaur has none and the Smoltaur fit covers the SmolLM2 models only.

Table 1: Core comparison of LLM-based cognitive and behavioural foundation models.
Model	Base LLM	Post-training	Training data	Domain
Centaur (11)	Llama-3.1-70B	SFT (masked CE on response tokens); rank-stabilised QLoRA (r=α=8), 4-bit quantised	Psych-101: 160 expts, 60,092 partic., 10.7M choices	Decision, memory, learning, planning
1021	Qwen2.5-7B-Instruct	SFT / Centaur-style SFT / GRPO compared; LoRA (r=α=32)	choices13k: 13,102 train / 1,462 test risky choice problems	Decision
Be.FM (95)	Llama-3.1-{8,70}B-Instruct	SFT with LoRA (all layers); 8-bit quantised2	AER: 2,703 papers.; MobLab: 68,779 subj., 82,057 obs.; Big Five: 17,667 subj.	Behavioural science, economic game, personality
Socrates (47)	Llama-3-8B-Instruct Qwen2.5-14B-Instruct	SFT / SFT + oracle reasoning traces / contrastive DPO compared; full fine-tuning	SocSci210: 210 TESS3 expts, 400,491 partic., 2.9M individual responses	Social sciences (economics, psychology, political science)
HumanLLM (55)	Qwen2.5-{3,7}B-Instruct Qwen3-8B4 Llama-3.1-8B-Instruct Phi-3-mini-128k-instruct	SFT (masked non-response tokens); full fine-tuning; 1:1 weight merge with base (LM-Cocktail)	Cognitive Genome: Reddit 2.8M, Twitter 673K, Blogger 368K, Amazon 1.7M; 1.2M train samples	Social intelligence (personalised behaviour)
GeCCo (66)	Llama-3.1-70B-Instruct DeepSeek-R1-Distill-Llama-3.1-70B Qwen2.5-72B-Instruct	No fine-tuning; in-context learning with iterative BIC-based refinement (10×5 runs)	Behavioural data from 4 cognitive domains (in-context, not for training)	Decision, learning, planning, working memory
Ours Llama-Centaur Qwentaur Smoltaur Olmotaur	Llama-3.2-{1,3}B Llama-3.1-8B Qwen3-{0.6,1.7,4,8,14}B-Base SmolLM2-{135,360}M SmolLM2-1.7B SmolLM3-3B-Base OLMo-2-0425-1B OLMo-3-1025-7B	SFT (masked CE on response tokens, Centaur-style); rank-stabilised LoRA (r=α∈{4,8,16,32,64})	Psych-101: 160 expts, 60,092 partic., 10.7M choices	Decision, memory, learning, planning
1No named model; methods (SFT v. RL) comparison only. 28-bit quantisation applies to 70B variant only; 8B is unquantised. 3TESS: NSF’s Time-sharing Experiments for the Social Sciences, a repository of peer-reviewed social science experiments conducted on nationally representative samples. 455 states “Qwen3-8B” without Instruct suffix; base/instruct status unspecified.

Figure 2: In-distribution versus out-of-distribution scaling at rank 16. (a) Psych-101, mean NLL over the 38 tasks with a reported cognitive model. (b) Psych-201-RT, all 18 held-out experiments; no cognitive baseline is published for Psych-201. Trend lines are fitted only within groups matched on generation and context window.

Table 2: Training hyperparameters for LLM-based cognitive and behavioural models. Adaptation strategy refers to the parameter-efficient or full fine-tuning method applied to the base LLM. All our variants share identical hyperparameters except for per-device batch size, gradient accumulation steps, and the set of adapter ranks trained. Entries marked “Not specified” indicate that the original paper did not report the corresponding detail.
Model	Adaptation	Optimiser	Epochs	Learning rate	Eff. batch size (PD×GA×d)	Scheduler
Centaur (11)	QLoRA (r=α=8), all linear layers	8-bit AdamW	11	5×10−5	1×32×1	Cosine (WU 100 steps)
102	LoRA r=α=32, all linear layers, dropout 0.05	AdamW	SFT: 6; RL: 3	SFT: 10−5; RL: 3×10−6	SFT: PD×8×1; RL: PD×8×4	SFT: fixed; RL: cosine
Be.FM (95)	LoRA, all layers	Not specified	3	10−4	1×8×d	Cosine (WU 0.1)
Socrates (47)	Full fine-tuning (no LoRA)	Not specified	1	SFT: 10−5; DPO: 10−6	PD×GA×8=256	Cosine (WU 0.05)
HumanLLM (55)	Full fine-tuning (no LoRA)	Not specified	3	5×10−6	PD×GA×8=64	Cosine (WU 0.5)
GeCCo (66)	N/A (no training)2	N/A	N/A	N/A	N/A	N/A
Qwentaur-0.6B	LoRA (r=α∈{4,8,16,32,64}), all linear layers	AdamW	1	5×10−5	2×16×1	Linear warmup (100 steps)
Qwentaur-1.7B	LoRA (r=α∈{8,16}), all linear layers	AdamW	1	5×10−5	1×32×1	Linear warmup (100 steps)
Qwentaur-4B	LoRA (r=α∈{4,8,16,32,64}), all linear layers	AdamW	1	5×10−5	1×32×1	Linear warmup (100 steps)
Qwentaur-8B	LoRA (r=α∈{4,8,16,32,64}), all linear layers	AdamW	1	5×10−5	1×32×1	Linear warmup (100 steps)
Qwentaur-14B	LoRA (r=α∈{4,16,64}), all linear layers	AdamW	1	5×10−5	1×32×1	Linear warmup (100 steps)
Llama-Centaur-1B	LoRA (r=α∈{4,8,16,32,64}), all linear layers	AdamW	1	5×10−5	2×16×1	Linear warmup (100 steps)
Llama-Centaur-3B	LoRA (r=α∈{4,8,16,32,64}), all linear layers	AdamW	1	5×10−5	1×32×1	Linear warmup (100 steps)
Llama-Centaur-8B	LoRA (r=α∈{4,8,16,32,64}), all linear layers	AdamW	1	5×10−5	1×32×1	Linear warmup (100 steps)
Smoltaur-0.1B	LoRA (r=α∈{4,8,16,32,64}), all linear layers	AdamW	1	5×10−5	8×4×1	Linear warmup (100 steps)
Smoltaur-0.4B	LoRA (r=α∈{4,8,16,32,64}), all linear layers	AdamW	1	5×10−5	4×8×1	Linear warmup (100 steps)
Smoltaur-1.7B	LoRA (r=α∈{4,8,16,32,64}), all linear layers	AdamW	1	5×10−5	2×16×1	Linear warmup (100 steps)
Smoltaur-3B	LoRA (r=α∈{4,8,16,32,64}), all linear layers	AdamW	1	5×10−5	1×32×1	Linear warmup (100 steps)
Olmotaur-1B	LoRA (r=α∈{4,8,16,32,64}), all linear layers	AdamW	1	5×10−5	2×16×1	Linear warmup (100 steps)
Olmotaur-7B	LoRA (r=α∈{4,8,16,32,64}), all linear layers	AdamW	1	5×10−5	1×32×1	Linear warmup (100 steps)
Abbreviations: PD = per-device batch size; GA = gradient accumulation steps; d = number of GPU devices; WU = warmup ratio. 1The published training script (https://github.com/marcelbinz/Llama-3.1-Centaur-70B/blob/main/scripts/cluster_train.sh) specifies 5 epochs. 11 report training for 1 epoch, suggesting early checkpoint selection. 2GeCCo (66) uses in-context learning with iterative BIC-based feedback over 10 sampling iterations × 5 independent runs. Models are fitted to held-out data with SciPy minimize (20 random restarts) and evaluated by BIC. All post-training methods use supervised fine-tuning unless otherwise noted; 102 additionally compare RL (GRPO with 12 candidate completions per step, max 1024 tokens; reward =1−\|oB−pB\| + format bonus up to 0.5, no standard-deviation normalisation); Socrates (47) additionally compares contrastive DPO (preference pairs constructed by varying the demographic persona under the same experimental condition and outcome question).

Figure 4: Structural ablation results across eight models and 27 experiments with well-defined response options. Five experiments with mixed or continuous formats are excluded. (a) Mean information retention R=(ln⁡k−ℒc)/(ln⁡k−ℒorig) by condition, averaged over experiments; R=1: no loss, R=0: chance. Circles: Llama-Centaur; squares: Qwentaur; colour intensity scales with model size; error bars: SEM. (b) Per-experiment fraction of learned information lost, δ=1−R, averaged across models. Columns sorted by history-only δ; colour strip indicates task type. Experiments with δ>1 under the history-only condition are omitted; see Figure 18 (Appendix F.2) for the complete set.

Table 3: Data pipeline and loss configuration for LLM-based cognitive and behavioural models. Precision refers to the numerical format used during training: quantised formats (4-bit NF4, 8-bit) apply to base model weights while LoRA adapters and forward/backward computation use bf16. Loss masking describes which tokens contribute to the training loss; masked approaches restrict gradient updates to human response tokens, avoiding optimisation on task instructions and context. Data format describes the input representation seen by the model during training. Data synthesis indicates whether and how any computer-assisted tools were used to construct or augment the training corpus.
Model	Precision	Weight decay	Loss masking	Data format	Data synthesis
Centaur (11)	4-bit NF4	0.01	Human response tokens only	NL trial-by-trial prompts (∼32K tokens)	Template-based prompt construction1
102	Not specified	Not specified	SFT: standard; Centaur-style. GRPO	JSON aggregated choice proportions per problem (empirical % rounded to nearest integer, e.g. {"A": 29, "B": 71})	Reformatted from choices13k empirical choice frequencies (problem-level rather than individual-participant-level prediction)
Be.FM (95)	8-bit (70B base, bitsandbytes); bf16 (8B)	Not specified	Standard (Alpaca template)	Alpaca template {instruction, input, output}	GPT-4o for research workflow extraction
Socrates (47)	Not specified	0.1	SFT: Response token only. DPO	{persona, stimuli, outcome, response}	o4-mini-high (dataset agent2); GPT-4o-mini (reasoning traces3)
HumanLLM (55)	Not specified	Not specified	Non-response positions masked	ShareGPT format	Llama-3.3-70B (extraction); GPT-4o (quality validation)
GeCCo (66)	N/A	N/A	N/A	NL prompt + Python function template	N/A
Ours	bf16	0.01	Human response tokens only (Centaur-style)	NL trial-by-trial prompts (∼32K tokens)	Psych-101 dataset (unmodified)
1Each experiment is converted into natural-language trial-by-trial prompts via author-written scripts that map structured experimental data (participant responses, stimuli, feedback) to verbalised narratives; for an example, see https://github.com/marcelbinz/Psych-201/blob/main/binz2022heuristics/generate_prompts.py. 2An LLM-based agent (o4-mini-high) parses raw TESS datasets into structured {persona, stimuli, response} tuples; see Appendix A of 47 for details. 3Given an experimental prompt and the corresponding human response, GPT-4o-mini generates reasoning traces explaining the human decision from a social scientist’s perspective; see Appendix E of 47.

Figure 5: Per-participant order variance across models and experiments. (a) THINGS odd-one-out (exchangeable; primary test). (b) Intertemporal choice (adaptive staircase; negative control). Left: violin and box plots; fine-tuned (solid) vs. base (hatched); y-axis scales differ. Right: ECDFs; solid = fine-tuned, dashed = base; steeper curves near zero indicate greater order invariance.

Table 4: Infrastructure, packages, and computational cost for LLM-based cognitive and behavioural models. Hardware refers to GPU resources used during training (or in-context generation for GeCCo). Training time reports wall-clock duration as stated in each paper; missing entries indicate the information was not reported. Our models were each trained for 1 epoch on Psych-101 on a single A100 80GB, with rank-stabilised LoRA at r=α∈{4,8,16,32,64}.
Model	Training Framework	Key packages	Hardware	Train time	Max seq. len.	Inference
Centaur (11)	unsloth	unsloth	1× A100 80GB	∼5 days	∼32,768	Not specified
102	Not specified	vLLM (inference)	RL: 4× H100; SFT: 1× A100	RL: ∼80 h; SFT: ∼5 h	1,024 (RL); 30 (SFT inf.)	T=0.7, top-p=0.95, top-k=0.5
Be.FM (95)	LlamaFactory	LlamaFactory, bitsandbytes	Not specified	Not specified	Not specified	Not specified
Socrates (47)	LlamaFactory	LlamaFactory	8× A100 80GB	4–24 h	4,096 (inf.)	T=0.6, top-p=0.9
HumanLLM (55)	LlamaFactory	LlamaFactory, DeepSpeed Zero, vLLM	8× A100 40GB	∼120 h (8B)	8,192	T=0.7
GeCCo (66)	N/A1	SciPy (minimize); Python exec()	4× A100 40GB	≤8 h per domain	N/A	T: 0.1–0.22
Qwentaur-0.6B	unsloth	unsloth	1× A100 80GB	4 h	∼32,768	—
Qwentaur-1.7B	unsloth	unsloth	1× A100 80GB	7 h	∼32,768	—
Qwentaur-4B	unsloth	unsloth	1× A100 80GB	18 h	∼32,768	—
Qwentaur-8B	unsloth	unsloth	1× A100 80GB	1 d	∼32,768	—
Qwentaur-14B	unsloth	unsloth	1× A100 80GB	1 d 15 h	∼32,768	—
Llama-Centaur-1B	unsloth	unsloth	1× A100 80GB	7 h	∼32,768	—
Llama-Centaur-3B	unsloth	unsloth	1× A100 80GB	12 h	∼32,768	—
Llama-Centaur-8B	unsloth	unsloth	1× A100 80GB	21 h	∼32,768	—
Smoltaur-0.1B	unsloth	unsloth	1× A100 80GB	3 h	∼8,192	—
Smoltaur-0.4B	unsloth	unsloth	1× A100 80GB	4 h	∼8,192	—
Smoltaur-1.7B	unsloth	unsloth	1× A100 80GB	7 h	∼8,192	—
Smoltaur-3B	unsloth	unsloth	1× A100 80GB	1 d 4 h	∼32,768	—
Olmotaur-1B	unsloth	unsloth	1× A100 80GB	5 h	∼4,096	—
Olmotaur-7B	unsloth	unsloth	1× A100 80GB	2 d 1 h	∼32,768	—
Abbreviations: T = sampling temperature; inf. = inference; d = days; h = hours. 1No training framework required; GeCCo generates cognitive models via in-context prompting. SciPy is used for parameter fitting of the generated models; Python exec() executes the LLM-generated code. 2Temperature varies by LLM: Llama 0.2, Qwen 0.15, DeepSeek-R1 0.1. All our models were trained on a single NVIDIA A100 80GB GPU. Training times scale approximately linearly with parameter count within each model family.

Figure 6: Adapter rank against model size on Psych-101. Mean negative log-likelihood over the 38 of 46 Psych-101 tasks for which 11 publish a domain-specific cognitive model baseline, against parameter count, with colour intensity encoding LoRA rank and right-hand panels showing one rank at a time. Trend lines and shaded envelopes are fitted only within groups matched on model generation and context window, so Olmotaur has none and the Smoltaur fit covers the SmolLM2 models only. Marker shape indicates context window (circle ≥ 32k, square < 32k); colour indicates family.

Table 5: Fraction of available information captured above chance, (ln⁡k−NLL)/ln⁡k, by task type for finetuned models (bf16). A value of 0 indicates chance-level performance; 1 indicates perfect prediction. Restricted to 34 experiments (of 46) with both a cognitive model baseline and a well-defined discrete response space (ln⁡k>0); 12 experiments are excluded. Finetuned families: Qwentaur, Llama-Centaur, Smoltaur, Olmotaur. Base columns: Q-8B/Q-14B (Qwen3), L-8B (Llama-3.1). Subscript r denotes our reproducing evaluation of the original Centaur model distributed by 11, evaluated under identical python library and CUDA versions as our small foundation models for fair comparison. Subscript p denotes values published by 11. 70Bp is shown for reference but excluded from best/second-best marking, as it was evaluated under different software conditions. Bold+underline marks the best model, underline the second-best.
	Qwentaur	Llama-Centaur	Smoltaur	Olmotaur	Base	11
Task type	0.6B	1.7B	4B	8B	14B	1B	3B	8B	0.1B	0.4B	1.7B	3B	1B	7B	Q-8B	Q-14B	L-8B	70Bp	70Br
Decision (8)	0.53	0.53	0.54	0.54	0.54	0.53	0.53	0.54	0.41	0.46	0.48	0.54	0.47	0.53	0.34	0.35	0.33	0.56	0.52
MDP (5)	0.28	0.30	0.31	0.31	0.32	0.28	0.30	0.31	0.18	0.20	0.26	0.30	0.20	0.30	0.15	0.17	0.19	0.32	0.31
Bandit (12)	0.48	0.48	0.49	0.49	0.50	0.46	0.48	0.49	0.36	0.41	0.44	0.49	0.42	0.48	0.35	0.36	0.33	0.46	0.45
Memory (4)	0.57	0.57	0.58	0.58	0.58	0.56	0.57	0.58	0.46	0.51	0.55	0.57	0.53	0.57	0.42	0.47	0.43	0.58	0.58
Misc. (2)	0.57	0.57	0.58	0.58	0.58	0.56	0.57	0.58	0.45	0.53	0.55	0.58	0.55	0.57	0.41	0.42	0.40	0.57	0.57
Sup. learn. (3)	0.30	0.29	0.31	0.31	0.31	0.29	0.31	0.31	0.18	0.23	0.25	0.30	0.20	0.31	0.24	0.24	0.23	0.30	0.30
Mean (34)	0.46	0.46	0.47	0.48	0.48	0.45	0.47	0.48	0.35	0.39	0.43	0.47	0.40	0.47	0.32	0.34	0.32	0.47	0.46

Figure 7: Adapter rank and training-set size by family. (Left) Mean NLL against LoRA rank at full data. (Right) Mean NLL against the fraction of Psych-101 used for training, at r=16. Rows are the four model families; the dotted line is the cognitive-model baseline and the diamonds are the reproduced and reported Centaur-70B values. Subsets are nested and experiment-stratified, so every fraction covers all 160 experiments and reducing data quantity does not reduce paradigm coverage.

Table 9: Fraction of available information captured above chance, (ln⁡k−NLL)/ln⁡k, by task type on Psych-201 (out-of-distribution). A value of 0 indicates chance-level performance; 1 indicates perfect prediction. Restricted to 15 experiments (of 18) with a well-defined discrete response space (ln⁡k>0); 3 experiments with continuous or mixed responses are excluded. Bold+underline marks the best model, underline the second-best.
	Qwentaur	Llama-Centaur	Smoltaur	Olmotaur	Base	11
Task type	0.6B	1.7B	4B	8B	14B	1B	3B	8B	0.1B	0.4B	1.7B	3B	1B	7B	Q-8B	Q-14B	L-8B	Centaur-70B
Decision (2)	0.30	0.32	0.35	0.34	0.37	0.27	0.31	0.33	0.20	0.21	0.24	0.34	0.21	0.31	0.28	0.29	0.28	0.33
MDP (2)	0.56	0.56	0.58	0.58	0.58	0.48	0.53	0.57	0.39	0.46	0.52	0.57	0.42	0.57	0.51	0.53	0.53	0.50
Bandit (4)	0.32	0.35	0.37	0.36	0.38	0.33	0.33	0.36	0.24	0.23	0.32	0.34	0.28	0.34	0.22	0.25	0.25	0.39
Misc. (7)	0.28	0.31	0.45	0.46	0.49	0.14	0.30	0.40	0.03	0.11	0.18	0.36	0.09	0.35	0.37	0.41	0.29	0.52
Mean (15)	0.33	0.35	0.43	0.43	0.45	0.26	0.34	0.40	0.16	0.20	0.27	0.38	0.20	0.37	0.34	0.37	0.31	0.46

Figure 8: Reproducibility of Centaur-70B (4-bit) evaluation across Psych-101. Reproduced vs. reported NLL for Centaur-70B (4-bit) across 46 Psych-101 experiments (r=0.994, mean Δ=+0.024, median Δ=+0.006).

Table 12: Impact of cognitive fine-tuning across different benchmarks. Each cell shows Δ = fine-tuned − base, where fine-tuned models are trained with LoRA r=16 on the full dataset. Significance is assessed with a two-sided z-test; group means use Stouffer’s method to combine per-task z-scores. EQ-Bench is on a separate scale and excluded from means.
	Qwentaur	Llama-Centaur
Benchmark	0.6B	1.7B	4B	8B	14B	1B	3B	8B
MetaBench
ARC	-0.03	-0.01	-0.01	+0.01	-0.01	+0.01	-0.05	+0.01
GSM8K	-0.25∗∗∗	-0.15∗∗∗	-0.05	-0.17∗∗∗	-0.12∗∗∗	-0.05∗	+0.00	-0.10∗
HellaSwag	-0.02	-0.09	-0.02	-0.03	-0.03	-0.04	-0.03	-0.06
MMLU	-0.09	-0.05	-0.06	+0.02	+0.02	-0.04	-0.05	+0.01
TruthfulQA	-0.01	+0.03	-0.05	+0.05	+0.03	-0.01	+0.04	+0.05
Winogrande	-0.02	-0.02	-0.02	+0.03	-0.02	-0.03	-0.04	-0.01
Mean (6)	-0.07∗∗∗	-0.05∗	-0.03	-0.01	-0.02	-0.03	-0.02	-0.02
Ethics
CM	+0.01	+0.00	+0.07∗∗∗	+0.01	+0.04∗∗∗	-0.02∗	-0.03∗∗	+0.00
Deontology	+0.02∗	-0.03∗	-0.02	+0.03∗	-0.04∗∗∗	+0.00	+0.05∗∗∗	+0.02∗
Justice	+0.06∗∗∗	+0.00	-0.01	+0.17∗∗∗	-0.08∗∗∗	+0.00	+0.05∗∗∗	+0.03∗
Utilitarian	+0.00	+0.02∗	+0.09∗∗∗	+0.03∗∗∗	+0.09∗∗∗	+0.00	-0.04∗∗∗	-0.01
Virtue	+0.56∗∗∗	+0.02∗	-0.01	+0.03∗∗∗	-0.05∗∗∗	-0.01	+0.20∗∗∗	-0.16∗∗∗
Mean (5)	+0.13∗∗∗	+0.00	+0.02∗∗∗	+0.05∗∗∗	-0.01	-0.01	+0.04∗∗∗	-0.02∗∗∗
Cog. & Lang.
LogiQA	+0.00	-0.01	-0.02	-0.02	-0.05	+0.00	-0.02	+0.00
PIQA	-0.02	-0.01	+0.01	+0.02	+0.01	-0.01	+0.00	+0.00
Social IQA	+0.00	+0.01	+0.03∗	+0.00	+0.01	+0.01	+0.01	+0.01
CoQA (F1)	-0.11∗∗∗	-0.03	-0.02	-0.01	+0.00	-0.07∗∗	-0.05∗	-0.03
LAMBADA (OAI)	+0.00	-0.01	+0.00	+0.00	+0.01	+0.01	+0.02∗	+0.01
LAMBADA (Std)	+0.02	+0.01	+0.03∗∗	+0.02∗	+0.02	+0.02∗	+0.03∗∗∗	+0.02∗
EQ-Bench	-44.7∗∗∗	+5.1	+5.9	-21.3∗∗∗	-0.3	+16.0∗∗∗	+22.9∗∗∗	+7.4
Mean (6)	-0.02	-0.01	+0.01	+0.00	+0.00	-0.01	+0.00	+0.00
ACP (Planning)
App (B)	-0.10	-0.10	-0.09	-0.02	-0.05	-0.29∗∗∗	+0.08	+0.05
Areach (B)	-0.04	-0.22∗∗∗	-0.24∗∗∗	-0.24∗∗∗	+0.01	-0.37∗∗∗	+0.14∗	+0.02
Just (B)	+0.05	-0.06	-0.06	-0.12∗	-0.01	-0.34∗∗∗	-0.13∗	-0.06
Land (B)	-0.12∗∗	-0.44∗∗∗	-0.15∗∗	-0.21∗∗∗	+0.10	-0.11∗∗	-0.13∗∗	-0.04

Figure 9: Specificity of cognitive fine-tuning: comparison with non-cognitive control models. Mean NLL on Psych-101 for cognitively fine-tuned models (Llama-Centaur, Qwentaur; LoRA r=16, full training data) and size-matched non-cognitive and cognitive controls (Hermes, Nemotron, Be.FM). Models are grouped by parameter count. Cognitively fine-tuned models consistently outperform non-cognitive controls at every scale, confirming that the improvement is specific to the behavioural signal in Psych-101 and not an artefact of fine-tuning per se.

Table 13: Prior ablation conditions expressed in the four-channel notation. ∙ = present, — = removed, Imin = reduced to a minimal action-space definition, ⋅~ = content replaced by generic placeholders with formatting preserved, ⊗ = replaced by an instruction that contradicts the task, ⋅π = trial order permuted.
Source	Condition	𝑰	𝑺	𝑭	𝑪	Closest condition in ours
11	Original	∙	∙	∙	∙	Original
94	No psychological task	Imin	—	—	∙	History-only
	Zero-shot prediction	∙	∙	—	—	no analogue
57	Instruction free	—	∙	∙	∙	Instruction-ablated
	Misleading instruction	⊗	∙	∙	∙	no analogue
	Context free	—	—	—	∙	Choice-only
Ours	Instruction-ablated	—	∙	∙	∙	—
	Content-masked	Imin	S~	F~	∙	new
	History-only	Imin	—	—	∙	—
	Choice-only	—	—	—	∙	—
	Order-permuted	∙	Sπ	Fπ	Cπ	new

Figure 10: Impact of cognitive fine-tuning on MetaBench performance (Δ = fine-tuned − base). Each bar shows the change in accuracy for a matched base–fine-tuned pair on six standard LM benchmarks (ARC, GSM8K, HellaSwag, MMLU, TruthfulQA, Winogrande) plus their mean. Error bars show pooled standard errors. Positive values (green region) indicate improvement; negative values (red region) indicate degradation. Colour intensity scales with model size within each family.

Table 15: Per-experiment NLL under sequential ablation conditions for Centaur-70B (11) on Psych-101 (in-distribution). Columns correspond to progressive prompt degradation: orig retains the full prompt; inst removes task instructions; cont additionally masks stimulus values and feedback; hist further removes trial structure, leaving only the response history. The ln⁡(k) column shows the random-guessing baseline where k is the number of per-trial response options. Experiments with mixed or continuous response formats have no well-defined k and are shown as – .
Experiment	Type	orig	inst	cont	hist	ln⁡(k)
Gardening task (28)	Decision	0.49	0.49	0.69	0.69	0.69
Columbia card task (32)	Decision	0.21	0.24	0.28	0.29	0.69
Experiential-symbolic task (34)	Decision	0.46	0.46	0.85	1.43	—
Multi-attribute DM (42)	Decision	0.06	0.08	0.74	0.74	0.69
Risky choice (50)	Decision	0.43	0.64	0.82	0.89	—
choices13k (63)	Decision	0.43	0.44	0.52	0.56	0.69
CPC18 (64)	Decision	0.35	0.35	0.38	0.41	0.69
Decisions from description (93)	Decision	0.59	0.61	0.69	0.68	0.69
Two-step task (48)	MDP	0.48	0.53	1.18	1.23	0.69
Two-step task (49)	MDP	0.53	0.54	1.15	1.16	0.69
Virtual subway network (84)	MDP	1.16	1.46	1.41	1.30	1.61
Multi-task RL (83)	MDP	0.57	0.66	0.77	0.82	1.10
Two-step task (104)	MDP	0.51	0.52	0.97	1.03	0.69
Drifting four-armed bandit (6)	Bandit	0.71	0.85	0.84	0.86	1.39
Horizon task (27)	Bandit	0.40	0.39	0.53	0.55	0.69
Two-armed bandit (35)	Bandit	0.30	0.36	0.39	0.46	0.69
Prob. instrumental learning (54)	Bandit	0.50	0.49	2.07	2.01	0.69
Horizon task (69)	Bandit	0.58	0.59	0.63	0.68	0.69
Structured bandit (72)	Bandit	0.64	0.68	0.84	1.00	2.08
Horizon task (77)	Bandit	0.35	0.36	0.53	0.61	0.69
Iowa gambling task (80)	Bandit	0.91	1.08	0.97	0.97	1.39
Horizon task (88)	Bandit	0.15	0.15	0.34	0.43	0.69
Horizon task (89)	Bandit	0.48	0.48	0.56	0.62	0.69
Spatially correlated MAB (91)	Bandit	1.82	1.95	2.45	2.62	3.40
Decisions from experience (93)	Bandit	0.47	0.95	0.54	0.57	—
Changing bandit (96)	Bandit	0.45	0.45	1.28	0.98	0.69
Cond. assoc. learning (19)	Memory	0.52	0.55	1.03	1.06	1.10
Shepard categorization (5)	Sup. learn.	0.54	0.58	0.70	0.70	0.69
Multiple-cue judgment (20)	Sup. learn.	1.14	1.19	1.94	1.94	2.20
Medin categorization (56)	Sup. learn.	0.50	0.58	0.79	0.90	—

Figure 11: Impact of cognitive fine-tuning on Ethics benchmark performance (Δ = fine-tuned − base). Each bar shows the change in accuracy for a matched base–fine-tuned pair across five ethical reasoning tasks (commonsense morality, deontology, justice, utilitarianism, virtue) plus their mean. Error bars show pooled standard errors. Positive values (green region) indicate improvement; negative values (red region) indicate degradation. Colour intensity scales with model size within each family.

Findings

Within the range of experiments used in training (in-distribution), models with 0.6 to 1 billion parameters matched the negative log-likelihood of a reproduced 70-billion-parameter Centaur-70B model, and all eight comparable models clustered within a narrow 0.028-nats band.
On entirely new experiments (out-of-distribution, Psych-201-RT), the same eight models spread across a much wider 0.244-nats range, with larger models clearly performing better.
Masking the specific content of stimuli and feedback destroyed 75.7% of the models' learned information and pushed performance below random-guessing level, while removing only the task instructions cost just 12.5%.
18 of 27 tested experiments (67%) showed strictly worsening performance as more prompt information was progressively removed.
On the THINGS similarity-judgement task, where trials are independent, shuffling trial order barely changed predictions, while on an intertemporal-choice task, where each trial depends on the previous response, shuffling order substantially changed predictions.

Figure 12: Impact of cognitive fine-tuning on cognitive and language benchmark performance (Δ = fine-tuned − base). Each bar shows the change in accuracy for a matched base–fine-tuned pair across six tasks (LogiQA, PIQA, Social IQA, CoQA, LAMBADA-OpenAI, LAMBADA-Standard) plus their mean. Error bars show pooled standard errors. Positive values (green region) indicate improvement; negative values (red region) indicate degradation. Colour intensity scales with model size within each family.

Where it can be used

Using small fine-tuned models instead of massive ones as a lower-cost tool for predicting human choice behaviour in research settings
Using these models as a noise-ceiling benchmark to gauge how much an existing psychological theory can still be improved for a specific experiment
Applying the channel-by-channel prompt removal method to check whether other behaviour-prediction models are actually using real task information

Figure 18: Complete per-experiment ablation heatmaps for 26 experiments with a well-defined chance baseline. Five additional experiments with mixed or continuous response formats are excluded throughout because no single k defines a chance baseline. The probabilistic instrumental learning task (54) is also omitted. Each cell shows the fraction of learned information lost, δ=(ℒc−ℒorig)/(ln⁡k−ℒorig), averaged across all eight models. Columns represent individual experiments sorted by δ under the history-only condition; rows represent ablation conditions. (Top) 20 experiments with mean δ≤1 under the history-only condition (i.e. performance remains at or above chance). (Bottom) 6 experiments where ablation degrades performance below chance (δ>1); note the separate colour scale. The colour strip below each panel indicates task type (see legend).

Limits and open work

Results were validated only on tasks resembling the 160 experiment types in Psych-101, so generalisation to genuinely novel experiment types remains unverified.
All models were fine-tuned only with low-rank adapters (LoRA); full fine-tuning at small scale and alternative architectures such as mixture-of-experts or state-space models were not tested.
The structural ablation tests covered only 27 experiments, and some task types, such as memory tasks where the stimulus itself is what must be recalled, had to be excluded because content and response cannot be separated.
The order-permutation test for exchangeability was run on only two contrasting experiments, one with independent trials and one with sequential dependency.
Cognitive fine-tuning was observed to reduce performance on general benchmarks like mathematical reasoning and formal planning, and its effect on ethical-reasoning benchmarks was inconsistent across models, so the authors did not draw firm conclusions there.

Why it matters

It had been assumed that predicting human behaviour with language models required massive, resource-hungry models like a 70-billion-parameter system, but this work shows much smaller models can match that performance within the range of tasks they were trained on. For psychology researchers and experiment designers, this points to a possible use for small fine-tuned models as a benchmark for how much of an experiment's behaviour is predictable at all, helping gauge how much room existing theories still have to improve.

Terms in this paper

Psych-101 · A dataset combining 10.7 million trial-level human choices from 160 psychology experiments
in-distribution / out-of-distribution · In-distribution means experiments similar to the training data; out-of-distribution means entirely new, unseen experiments
LoRA (adapter) · A lightweight fine-tuning method that trains only a small added component instead of retraining the whole model
noise ceiling · The theoretical best possible prediction accuracy on a dataset, limited by the unavoidable randomness in human behaviour itself
exchangeability · A property where trials are independent of one another, so their order carries no predictive information

Original abstract (English)

Large language models fine-tuned on human behavioural data have emerged as general-purpose cognitive proxies, but the scale this requires, and whether these models process task structure or exploit statistical shortcuts, remain open questions. We train fourteen models from 135M to 14B parameters across four architecture families on Psych-101, a dataset of 10.7 million trial-level choices from 160 experiments. In-distribution, scale barely matters. The models fall within a narrow band, as though against a ceiling, and 0.6B to 1B parameters suffice to match a 70B baseline on held-out participants. Out-of-distribution, that band opens into a markedly steeper scaling gradient, with larger models clearly advantaged in generalisation to novel task structure. To determine what information these models use, we run two diagnostics. We progressively strip four prompt channels -- task instructions, experimental stimuli, outcome feedback, and choice history -- across 27 experiments, and permute trial order. Masking the content of stimuli and feedback destroys 75.7% of learned information and pushes models below chance, demonstrating that choice history alone does not account for performance. Permutation reveals invariance on tasks with independent trials but sensitivity where trial order is determined by prior responses. Small cognitively fine-tuned models therefore show promise as noise ceiling estimators for psychological experiments, though their scope remains bounded by the paradigms seen in training.

Authors · Nick Oh

Read on arXiv

Latest papers

All papers →

Latest from METAL MEDIA

Figures: Nick Oh et al., arXiv:2608.05224, CC BY 4.0