Partner-Specific Affective Precision in Social Active Inference

Harshil Shah; Andrew Pashea

cs.AISep 21, 2026

Partner-Specific Affective Precision in Social Active Inference

Authors: Harshil Shah, Andrew Pashea

Organizations: Mission San Jose High School, Fremont, CA, USA · Division of the Social Sciences, University of Chicago, Chicago, IL, USA

Abstract

In multi-agent social settings, model reliability varies across relationships. Beyond inferring what others will do, an agent must calibrate how confidently those inferences should guide policy selection for each relationship. An agent may maintain a well-validated model of one partner, a fragile model of another, and a model under revision for a third; collapsing these into a single confidence estimate loses information relevant to policy selection. We therefore formalize affective precision as a relationship-specific metacognitive estimate of confidence in the current partner model. Each partner's behavioral evidence updates a local confidence estimate that modulates policy precision during selection, regulating how strongly current beliefs are expressed in policy rather than changing the content of those beliefs. Simulations in a multi-partner graded trust game show that partner-local affective precision influences behavior primarily through policy commitment rather than direct improvement of partner-state inference. Because the mechanism tracks partner-response predictability rather than realized payoff, greater confidence produces sharper policy commitment without necessarily producing higher rewards. Under abrupt shifts in social behavior, confidence accumulated from previously reliable predictions can remain behaviorally active after the relationship changes, showing that confidence revision can lag behind social change. Finally, varying precision gain and priors produce distinct trust-calibration dynamics, showing how confidence accumulation and revision depend on model parameters. Together, these results show how relationship-specific affective precision can distinguish social prediction from social policy commitment.

Figures & tables

Figure 1: Partner-local versus shared precision in the locality probe (30 seeds per condition). Left and middle: absolute active-encounter partial correlations of precision with surprisal and payoff, controlling for encounter count and the alternative signal. Right: cumulative payoff. Error bars: 95% CIs.

Figure 2: Tracked-only ablation in the open graded partner-choice regime. Bars show means across 30 seeds for within-episode

\bar{\beta}_{k}

range, policy entropy, and cumulative payoff (95% CIs). Partner-local precision lowers entropy while tracker movement remains comparable. Paired contrasts are reported in the text.

Figure 3: Abrupt betrayal in the partner-choice regime. Left and right panels compare partner-local precision with no-affect; the middle panel compares partner-local precision with tracked-only. Dashed vertical lines mark partner 0’s round-31 stance switch. Lines show means within 10-round bins across 30 seeds; shading: 95% CIs.

Appendix figures & tables12 assets

Supplementary material from the paper’s appendix.

Appendix

Partner type	Trusting	Neutral	Hostile
Cooperator	0.95	0.80	0.55
Reciprocator	0.90	0.70	0.30
Exploiter	0.70	0.35	0.10
Random	0.60	0.50	0.35

Appendix

Table 1: Cooperation likelihood

P(\text{cooperate}\mid s^{\text{type}},s^{\text{stance}})

High investment / trust-building endpoint $B^{\mathrm{high}}$
To $\setminus$ From	Trusting	Neutral	Hostile
Trusting	0.90	0.30	0.05
Neutral	0.10	0.60	0.35
Hostile	0.00	0.10	0.60
Low investment / withholding endpoint $B^{\mathrm{low}}$
To $\setminus$ From	Trusting	Neutral	Hostile

Appendix

Table 2: Endpoint stance transition dynamics. The focal agent’s graded investment interpolates between

B^{\mathrm{high}}

and

B^{\mathrm{low}}

according to Eq. ( 6 ).

Condition	$\beta$ tracker and $\gamma_{k}$	Precision scope
No-affect	off; $\gamma_{k}=\gamma_{\text{base}}$ fixed	—
Tracked-only	$q(\beta_{k})$ updated; $\gamma_{k}=\gamma_{\text{base}}$ fixed	per-partner
Partner-local	$q(\beta_{k})$ updated; $\gamma_{k}$ from $q(\beta_{k})$	per-partner
Shared- $\beta$	pooled $q(\beta)$ updated; $\gamma_{k}$ from shared $\beta$	pooled tracker; separate beliefs

Appendix

Table 3: Precision ablation conditions

Condition	Purpose
Open partner choice	Tracked-only deployment contrast and descriptive partner-type allocation readout from the same trajectories
Locality probe	Precision–surprisal vs. precision–payoff correlations under local/shared $\beta$
Abrupt betrayal	Episode-level policy entropy, payoff, and joint type–stance accuracy, with post-switch partner-selection, precision, and entropy trajectories
Profile variation	$\alpha$ /prior manipulations for trust-calibration profiles
Forgiveness	Reengagement and $\beta$ recovery after repair

Appendix

Table 4: Simulation condition families and purpose

Experiment family	Setting	Rounds	Seeds
Open deployment / allocation	graded partner choice; shared trajectories for Sections 3.2 – 3.3	200	30
Locality probe	graded partner choice	200	30
Abrupt betrayal	scheduled partner-choice betrayal	120	30
Profile program	sweeps, factorial, forgiveness	200	20

Appendix

Table 5: Representative episode lengths and replication counts by experiment family

Metric	Definition
Total payoff	Cumulative focal payoff over episode/window
Policy entropy	Natural-log Shannon entropy of $q(\pi)$ , averaged over rounds/seeds
Joint type–stance accuracy	Mean simultaneous correctness of inferred partner type and stance
Selection share	Fraction of rounds each partner is chosen
Selection concentration	Gini coefficient over partner-selection distribution
Mean within-episode $\bar{\beta}_{k}$ range	Average within-episode max–min range of posterior mean $\bar{\beta}_{k}$ , averaged across seeds and episodes

Appendix

Table 6: Metric glossary

$\alpha$	Mean within-episode $\bar{\beta}_{k}$ range
0.05	0.097
0.1	0.126
0.3	0.195
0.5	0.219
1.0	0.277
2.0	0.402

Appendix

Table 7: Mean within-episode

\bar{\beta}_{k}

range by precision-gain

\alpha

(betrayal regime)

Figure 4: Precision-gain sweep across open graded and betrayal environments (20 seeds per

\alpha

/environment). Points show means with 95% CIs, excluding undefined recovery times. The entropy panel pools both environments (40 runs per

\alpha

); early, mid, and late denote rounds 1–50, 51–150, and 151–200. All horizontal axes show gain

\alpha

on a logarithmic scale. Other metric definitions appear in Table 6 .

Profile	Payoff	Recovery (rounds)	Gini	Mean $\bar{\beta}_{k}$ range	Trust asymmetry
Anxious-reactive	2285.8	25.1	0.626	0.459	14.87
Hypervigilant	2230.7	21.8	0.609	0.556	6.97
Naive-stubborn	2218.0	15.5	0.567	0.173	4.87
Avoidant-rigid	2166.0	13.8	0.567	0.302	1.47
Default	2244.2	31.4	0.643	0.427	7.79

Appendix

Table 8: Behavioral metrics by profile in the betrayal environment; the default agent is shown for reference. Values are means from 20 seeds per profile, excluding undefined recovery times. The

\bar{\beta}_{k}

range is the within-episode max–min range, averaged over partners and seeds.

Figure 5: Gain-prior profile signatures in the betrayal environment. Printed cell values are raw means; color scales each metric separately from the lowest to the highest profile mean. Thus, colors compare profiles within a column, not magnitudes between metrics. Table 8 gives the numerical summary.

Variant	Reengagement	Payoff recovery	Latency (rounds)	Mean $\bar{\beta}_{k}$ range
cautious-high- $\alpha$	0.529	1.043	14.5	0.607
cautious-low- $\alpha$	0.684	1.063	6.7	0.301
default	0.584	1.021	10.3	0.485
naive-high- $\alpha$	0.577	0.989	3.5	0.604
naive-low- $\alpha$	0.613	1.010	7.9	0.190
no-affect	0.586	1.042	12.4	–

Appendix

Table 9: Forgiveness and trust-repair metrics (20 seeds per profile). Reengagement is the post-repair partner-0 selection rate; payoff recovery uses the windows in Figure 6 . Latency averages only seeds that reengage (16–20 per profile). The dash denotes the absent no-affect

\beta

tracker.

Figure 6: Forgiveness and trust repair (means across 20 seeds per profile). A: partner-0 posterior mean

\beta_{k}

; shading marks betrayal (rounds 81–120), followed by repair. Lower

\beta_{k}

indicates higher policy precision; no-affect has no tracker. B: partner-0 selection during repair. C: payoff in rounds 151–200 relative to rounds 50–80; the dashed line at 1 marks baseline recovery.

Explore similar work

Learning Whom to Trust : Decision-Generated Credibility in Social Learning
Aug 25, 2026Gabriel Bontemps, Abhishek BanerjeeComputational Social Science
GREDEG-CNRS, Université Côte d’Azur, Nice, France · Blizard Institute, Queen Mary University of London, London, United Kingdom of Great Britain and Northern Ireland (the) · Adaptive Decisions Laboratory, Department of Pharmacology, University of Oxford, Oxford, United Kingdom of Great Britain and Northern Ireland (the)
Trust Between AI Agents: Measuring Formation, Breakage, and Recovery, with Implications for Governing Multi-Agent Systems
Jun 12, 2026Yujiao ChenAI Agent Evaluation Multi-Agent Systems
Massachusetts Institute of Technology Cambridge, MA 02139
Mind or Message? Auditing Theory of Mind in Multi-Agent Social Simulation
Sep 21, 2026Cong Li, Cheng Chen, Thomas Fung +2Theory of Mind LLM Agent Evaluation
DailyAdvance

Aug 25, 2026, cs.NEJ/K move · Enter open · S save

Learning Whom to Trust : Decision-Generated Credibility in Social Learning

Gabriel Bontemps, Abhishek Banerjee

GREDEG-CNRS, Université Côte d’Azur, Nice, France · Blizard Institute, Queen Mary University of London, London, United Kingdom of Great Britain and Northern Ireland (the) · Adaptive Decisions Laboratory, Department of Pharmacology, University of Oxford, Oxford, United Kingdom of Great Britain and Northern Ireland (the)

Computational Social Science

Open paper AskPDF arXiv

Social interaction can improve collective learning but also amplify early mistakes. We study this tension when the credibility of social information is generated by the sender's own decision process rather than fixed ex ante. Reinforcement-learning agents make binary choices through a drift--diffusion process that jointly determines choice, decision time, and confidence; decision confidence then becomes social credibility by weighting anticipatory influence and retrospective social learning. Under balanced community exposure, the anticipatory field admits an exact quotient representation. Its local Jacobian is a scalar decision-sensitivity term multiplying the community-coupling matrix, which yields a common-mode amplification threshold and an analytical role for cross-community permeability in damping relative community differences. Monte Carlo experiments show the corresponding non-monotone performance pattern: moderate transmission accelerates correction, whereas strong transmission can lock populations into wrong consensus; low permeability instead sustains disagreement. Ablations reveal a dual role for confidence: credibility-sensitive transmission amplifies social error, while confidence-dependent private learning stabilises it. The model yields testable predictions linking sender confidence to receiver behaviour conditional on accuracy.

Abstract

Explore similar work

Learning Whom to Trust : Decision-Generated Credibility in Social Learning

Trust Between AI Agents: Measuring Formation, Breakage, and Recovery, with Implications for Governing Multi-Agent Systems

Mind or Message? Auditing Theory of Mind in Multi-Agent Social Simulation