MOTIVATION LETTER The CCT model is a tripartite pharmacological framework for reward-memory encoding prevention in addiction, built as a coupled three-axis ODE system with Bayesian MCMC calibration. The model has confirmed all five pre-registered hypotheses (H1-H5) with posterior super-additivity of 13-22 percentage points across model versions, and three sole-authored preprints are under review at IART, PNPBP, and NBR. This work addresses a global health burden that disproportionately affects low- and middle-income countries, including Nigeria, where addiction treatment infrastructure remains limited. Open Philanthropy's long-termist focus areas include global health and AI risk. The CCT model speaks to both. On global health, the framework identifies a pharmacological mechanism to prevent the consolidation of reward memories, which is the core neurobiological process underlying addiction relapse. On AI risk, the model's architecture mirrors reward hacking and corrigibility problems in AI alignment: the same dopaminergic reward prediction error dynamics that drive maladaptive learning in addiction are formal analogues of reward misspecification in learning systems. The computational methods used, Bayesian calibration with DEMetropolisZ sampling and dynamical systems analysis, transfer directly to AI safety research on reward modeling. The programme's selection criteria emphasize cost-effectiveness and urgency. The CCT model is cost-effective because it is a computational research program with no wet-lab overhead; the entire calibration pipeline runs on standard HPC infrastructure. The urgency stems from the scale of the problem: opioid and stimulant use disorders are rising across Africa, and existing pharmacological interventions have limited efficacy precisely because they do not address the memory consolidation mechanism this model targets. The applicant profile includes several independent research lines, but the CCT model is the strongest fit for this programme. The cardiotoxicity topology study and TOPOLOGIX address drug resistance prediction, which is valuable but less aligned with Open Philanthropy's stated focus. The ergofluids work is behind a real-data validation gate and is explicitly methods-validation research. The CCT model is the only line that directly connects a global health burden to a formal framework with AI safety relevance, and it has the strongest evidential track record: all pre-registered hypotheses confirmed, three peer-reviewed preprints in review, and a co-authored paper under review at Alcohol (Elsevier). The work is independent and self-funded to date. Funding from Open Philanthropy would support the next phase: extending the model to opioid-specific circuits, validating against human neuroimaging data, and preparing the framework for clinical trial design. The M.Sc. in Digital Health at Hasso Plattner Institute, beginning Winter Semester 2026/27, provides institutional grounding for this next phase. RESEARCH STATEMENT The Conjunctive Consolidation Threshold (CCT) model addresses a specific gap in addiction neuroscience: no existing framework explains why some reward experiences consolidate into persistent drug-seeking memories while others do not. The model proposes a tripartite threshold mechanism requiring simultaneous activation of three coupled axes: dopaminergic reward prediction error, NMDAR-dependent long-term potentiation, and affective contrast. When all three exceed a conjunctive threshold, the reward memory is encoded; when any axis falls below threshold, encoding fails. The model is implemented as a system of coupled ordinary differential equations solved with RK45 integration. The parameter space includes 14 free parameters, calibrated using Bayesian MCMC with the DEMetropolisZ sampler in PyMC. Prior distributions were elicited from a systematic screen of 1,847 records from the addiction neuroscience literature. All five pre-registered hypotheses (H1-H5) were confirmed. The key finding is posterior super-additivity: the combined effect of the three axes exceeds the sum of individual effects by 13-22 percentage points across model versions. This super-additivity is the formal signature of a conjunctive threshold, and it has direct pharmacological implications: interventions targeting any single axis are predicted to be insufficient, while interventions that push any axis below threshold should prevent memory encoding. Three sole-authored preprints describing the model, its calibration, and its pharmacological predictions are under review at IART, PNPBP, and NBR. A co-authored paper is under review at Alcohol (Elsevier). The model has been extended into neurocascade, a receptor-to-behavior simulation engine that couples pharmacokinetics to receptor binding to Wilson-Cowan circuit dynamics to behavioral readouts. Three literature-calibrated receptor systems (mu-opioid, D2 dopamine, GABA-A) are implemented with 62 of 62 tests passing. The relevance to Open Philanthropy's mission is twofold. First, addiction is a global health crisis with long-term consequences for human welfare and economic development, particularly in LMICs. The CCT model provides a mechanistic target for pharmacological intervention that could prevent addiction before it consolidates, rather than treating it after the fact. Second, the model's architecture is a formal analogue of reward hacking in AI systems. The dopaminergic reward prediction error signal that drives maladaptive learning in addiction is the same signal class that drives reward misspecification in reinforcement learning agents. The conjunctive threshold mechanism offers a potential corrigibility intervention: a formal method for preventing the consolidation of misspecified reward functions. The computational methods are directly transferable to AI safety research. Bayesian calibration of dynamical systems models is the standard toolkit for uncertainty quantification in safety-critical systems. The model's treatment of multiple evidence streams, where disagreement between streams is preserved rather than averaged away, is a formal approach to the problem of conflicting objectives in AI alignment. The next phase of research has three components. First, extend the model to opioid-specific circuits using the mu-opioid receptor system already implemented in neurocascade. Second, validate the model's predictions against human neuroimaging data on reward memory consolidation. Third, translate the conjunctive threshold framework into a testable clinical protocol for pharmacological prevention of addiction. This phase requires computational infrastructure, access to neuroimaging datasets, and collaboration with clinical researchers. Funding from Open Philanthropy would support all three components. EDITOR NOTES - Research line selected: CCT model. This is the only line in the profile that directly connects a global health burden (addiction) to a formal framework with AI safety relevance (reward hacking, corrigibility), matching Open Philanthropy's stated focus areas. The cardiotoxicity/TOPOLOGIX work is drug resistance prediction, which is not aligned with the programme's mission. ergofluids is behind a real-data validation gate and is explicitly methods-validation research, not suitable for this grant. - Eligibility risk: The programme's stated eligibility is for grantees affected by the FTXFF collapse. The applicant profile does not indicate any FTXFF connection. This is a material risk that must be flagged before submission. The application should be submitted only if the applicant can verify eligibility or if the programme has expanded eligibility since the profile was compiled. - Verification needed: The three preprints under review (IART, PNPBP, NBR) and the co-authored paper under review at Alcohol (Elsevier) need current status confirmation. The claim that the CCT model is the only research line with AI safety relevance should be checked against the programme's current funding priorities, which may have shifted. - Gap to fill: The applicant must insert specific details about how the CCT model's next phase (opioid-specific circuits, neuroimaging validation, clinical protocol) would be executed within the funding period, including timeline, collaborators, and institutional support from Hasso Plattner Institute. - Tone check: The motivation letter opens with the research, not with self-introduction, per formatting rules. All claims are specific and quantified. No AI-slop phrases used. The letter does not claim validated IP or product-market fit for any venture; the CCT model is described as a research framework with confirmed pre-registered hypotheses, which is accurate per the profile. CHECKLIST - [ ] Verify eligibility for Open Philanthropy Grant Program (FTXFF-affected grantee status) - [ ] Confirm current review status of three preprints (IART, PNPBP, NBR) - [ ] Confirm current review status of co-authored paper at Alcohol (Elsevier) - [ ] Insert specific timeline and budget for next-phase CCT research (opioid circuits, neuroimaging validation, clinical protocol) - [ ] Confirm Hasso Plattner Institute institutional support and M.Sc. enrollment details - [ ] Verify ORCID (0009-0001-9272-6735) and GitHub (github.com/AmunRaPtah) links are active - [ ] Prepare CV in Open Philanthropy's preferred format - [ ] Submit application via the URL in the programme profile - [ ] Confirm no additional submission page content was required beyond this document