iiinova — where your digital self keeps evolving

Big picture: human problem solving is more than search & planning in an MDP—naturalistic challenges may change our strategies fundamentally! We need lab experiments in dialogue with field studies, and theories that treat past solutions and analogies as central—not accessory—sources of intelligence.

020 2

Process data told the same story. In itinerary planning, people prioritized cost labels along cached routes, but revealed them in less orderly sequences—consistent with checking whole routes and estimating order-agnostic total costs, as opposed to sequentially expanding a biased search frontier.

100 1

Across experiments spanning navigation, alchemy, and itinerary planning, people shifted toward solution-driven modes amid high representational costs. They reused cached solutions even when suboptimal, but not blindly: they remained sensitive to which cached solution worked best in each situation.

100 1

This representational challenge is salient in everyday life. Field studies have long argued that humans are predominantly solution-driven, but their reliance on expert introspection makes it hard to separate this from blind habit or biased search. We now test these ideas in controlled lab paradigms.

100 1

Importantly, we characterize solution-driven problem solving as a response to representational difficulty, not merely downstream planning costs. By starting from candidate solutions, people inherit a problem frame in advance, avoiding costly search over the huge space of possible representations.

100 1

How do humans find good representations amid this challenge? We propose a *solution-driven* account that reverses the classical order: people retrieve cached solutions first, then represent only the features needed to evaluate and compare them. Solutions come before representations, not after.

100 1

Classic theories cast human problem-solving as search within a problem representation. When none is given, we must search for good representations first. But evaluating a representation requires asking if the plans derived from it would work—often forcing us to represent the full problem after all.

Liu, Shuze @liushuze.bsky.social · 18/06/2026

Do people solve hard problems by first building a mental model, then searching for the best solution? With @gershbrain.bsky.social, our new manuscript suggests a different route: people often start by retrieving cached solutions, and scaffold representations around them. osf.io/preprints/ps...

14319 43

Join us at both conferences this summer! Our #CogSci2026 workshop features a series of invited talks on problem representations and abstractions. Our #CCN2026 community event opens a broader discussion on naturalistic problem solving: frameworks, tasks, methods, and where the field should go next.

0195 19

The study's message extends beyond the lab. As we increasingly face externally provided, prefitted models—folk theories, political narratives, or purported expert wisdom—in modern life, a tendency to overprefer complex predictive models may profoundly influence our individual and social well-being!

040 4

We found that humans are undersensitive to overfitting: they overrely on training loss as a proxy for test loss, and penalize other Efron terms suboptimally. Typical decision noise worsens the picture: people should have chosen the simple predictive model most of the time in our task, yet few do so.

So, do people intuitively understand overfitting? We generated noisy scatterplot data and asked humans to choose between predictive models fitted to the data. We compared humans to Efron's decomp, which predicts test loss from training loss, model complexity, training sample size, and data noise.

There is abundant literature from statistics and ML on predictive model overfitting. When data is scarce and noisy, an underparameterized model may be preferred over flexible models that reflect the ground truth, as their fitted params would likely capture training data noise and generalize poorly.

Human model selection studies often focus on generative model families, exemplified by the Bayesian Occam's razor that integrates over parameter values. Yet outside the lab, we often rely on prefitted models from others, and success is judged not by generative truth but by predictive accuracy.

Thanks Sam! Main takeaways: 1) Ground-truth vs. predictive model selection differ under noisy and scarce data—for prediction, oversimplified models may work better in avoiding overfitting. 2) When humans decide between externally provided, prefitted predictive models, they're undersensitive to 1).

1263 26

Overall, by jointly studying constraints on action sets and policy complexity, we provide a general picture of how humans adjust the two together. The results reveal lossy problem simplification beyond typical modeling assumptions, which may be an integral part of naturalistic human decision-making!

030 3

In a large-action-space contextual bandit experiment, we found that humans exploit the above interplay, enlarging action sets alongside policy complexity to mitigate suboptimality. Under time limits, they remain near-optimal for their chosen action set, indicating spontaneous problem simplification.

Beyond explaining past data, our framework prescribes a complex interaction. Enlarging the action set size uncaps policy complexity and enables greater reward. It also mitigates the increase in suboptimality following policy complexity increments, boosting the reward-efficiency of cognitive effort!

Using rate-distortion theory, we assess suboptimalities incurred by smaller action consideration set sizes at various levels of policy state-dependence. We rationalize empirical signatures of human option generation as adaptations to joint limitations on action set size and policy complexity.

In real-life decisions, vast action spaces often preclude our exhaustive consideration. Furthermore, cognitive constraints limit the state-dependence of policies that map world states to the actions considered. We build a resource-rational framework unifying both ecologically relevant constraints!

130 3

Liu, Shuze @liushuze.bsky.social · 28/07/2025

@gershbrain.bsky.social and I have a new paper in PLOS Comp Bio! We study how two cognitive constraints—action consideration set size & policy complexity—interact in context-dependent decision making, and how humans exploit their synergy to reduce behavioral suboptimality. osf.io/preprints/ps...

osf.io

OSF

12611 26

Finally, a huge thanks to my mentors @gershbrain.bsky.social and Bilal Bari for their support, insight, and encouragement! For those interested, we also have an earlier JEP:G paper, which inspired the mental cost modeling in our CogSci submission: gershmanlab.com/pubs/Liu25.pdf

020 2

We found that humans meta-reason their policy complexity according to both time and mental costs, exhibiting consistently supralinear mental cost functions across tasks. This overturns common assumptions, and supports the construct validity of info-theoretic measures as a domain-general mental cost!

To address this literature gap, we designed a series of contextual bandit experiments---addressing speed-accuracy tradeoffs, working memory set size manipulations, and reward magnitudes---to stress-test the mutual information formulation of mental costs and look for consistent relationships.

While rational inattention & policy compression have formulated mental costs via mutual information, there remains disagreement on whether it induces a capacity limit or a linear cost. Time costs further confound the picture, as they also incentivize low policy complexity to reduce decision time.

It is well known that context sensitivity incurs mental costs. However, it is unclear which domain-general cognitive resources underlie mental costs, and whether existing cost formulations have construct validity--which likely requires them to scale robustly with the assumed resource substrate.

Weiji Ma @weijima.bsky.social · 10/06/2025

Looking forward to sharing our work at #cogsci2025! We aim at getting one step closer to a domain-general formulation of mental costs via policy compression. Come see my presentation at "Talks 35: Reasoning". It is scheduled at 16:44 PST, August 1 at Nob Hill C! gershmanlab.com/pubs/LiuGers...

gershmanlab.com

1295 29

Reposted by Liu, Shuze

This is an unusual one: the project started in 2009, with Trevor Holland collecting data in an auditory-visual integration paradigm with many conditions within subjects. But modeling it remained challenging over the years until Luigi Acerbi and Shuze Liu took it on. osf.io/preprints/ps...

osf.io

OSF

193 9

Overall, we stress-test the Bayesian account of multisensory perception by systematically traversing its full modeling space. Human behavior remains well-explained, but only under specific, often overlooked assumptions. A richer picture emerges when we let data guide our modeling assumptions!

020 2

Beyond core inference, other complex perceptual factors play a role too: 1) Sensory noise increases in multisensory trials, likely due to divided attention; 2) Auditory observations are stretched according to the visual range, suggesting spontaneous cross-modal recalibration in humans.

We found that key model choices drastically affect fits and better explain the human central tendency: 1) Human priors are non-Gaussian; 2) Sensory noise is heteroskedastic, dipping centrally and plateauing peripherally; 3) Both model averaging (optimal) and probability matching fit behavior well.

We explore modeling choices in Bayesian cue integration—priors, sensory noise functions, and causal inference strategies—using a data-driven, semiparametric approach. With promising candidates identified, we enumerate them in a combinatorial model space and test them via model comparison.

140 4

@weijima.bsky.social, @lacerbi.bsky.social, Trevor, and I have a new preprint on Bayesian models of multisensory perception! We systematically inspect the underexplored degrees of freedom in Bayesian models, squeezing out their best capability in capturing human behavior. osf.io/preprints/ps...

1141 14

Overall, our study: 1) Connects seemingly disparate cognitive measures: state-dependence v. RT, goal-directed v. habitual behavior; 2) Prescribes task-general insight via normative principles; 3) Highlights the utility of incorporating multiple resource formulations in resource-rational studies!

010 1

Across three experiments, humans adaptively adjusted policy complexity in the predicted directions (though with a leftward bias, which we model via memory costs in our CogSci paper). LBA modeling revealed that policy-compression-style perseveration had manifested strongly in participant behavior.

Given policy complexity-RT relations, we can derive policy complexity levels that maximize reward over time. This generates predictions across various task manipulations, including ITIs, reward regularities, and set sizes (reward magnitudes forthcoming in CogSci 2025), which we test in this paper.

Policy compression applies rate-distortion theory to action selection, specifying the attainable reward at every policy state-dependence/complexity level. It prescribes a linear relationship b/w policy complexity and RTs, and rationalizes action perserveration as optimal usage of limited resources.

Tailoring actions to states taxes cognitive resources. Two prominent resource formulations are time and memory, studied in speed-accuracy tradeoffs and set-size effects. We unify them under policy compression, prescribing how humans should adaptively adjust the state-dependence of their policies.

@lucylai.bsky.social, @gershbrain.bsky.social, Bilal Bari, and I have a new paper out in JEP:G! We study the time and memory costs of policy compression—a resource-rational framework for decision making, focusing on how state-dependent our policies ought to be. psycnet.apa.org/record/2026-...

psycnet.apa.org

APA PsycNet

1131 13

By offering an initial characterization of how people select and switch among assets, we aim to lay the groundwork for understanding human decisions in temporally extended contexts, where uncertain future potential—as opposed to immediate payoff—plays a central role in our current reasoning. (6/6)

010 1

Across three experiments spanning asset selection and switching scenarios, varying time horizons, and different statistical properties of asset trajectories, we show that the Bayesian probabilistic forecasting model consistently outperforms myopic heuristics inspired by existing frameworks. (5/6)

100 1

Here we develop a normative model: humans use Bayesian function learning to extrapolate asset performance trajectories, making decisions based on forecasted outcomes at a task-specific time horizon. Prior knowledge informs these inferences, enabling comparisons between current and new options. (4/6)

Previous frameworks—foraging and study-time allocation—offer partial insights, but they typically rely on myopic assumptions, overlooking potential asset improvement. While models of human forecasting exist, it remains unclear how forecasts shape human decisions on when to commit or switch. (3/6)

100 1

When should we persist, and when should we switch? Abandoning an ongoing investment often means forfeiting the time spent and starting over from scratch. One may avoid the dilemma by choosing a good asset to begin with. But how do we make such decisions at the beginning or halfway through? (2/6)