Evidence behind the
Trident G protocol

Cognitive-control readiness and Zone Pulse

The Majority Function Task with masking estimates how much task-relevant perceptual information can be brought under cognitive control under time pressure. Wu et al. modelled this capacity as an information-processing limit, while He et al. developed an adaptive assessment method. A seven-day training study subsequently reported gains in cognitive-control capacity and selective transfer to some attention, learning, and neural processing measures, with effects varying by task similarity (Wu et al., 2016), (He et al., 2022), (Zhang et al., 2024).

Meaning for Trident G: Zone Pulse has a defensible basis as a readiness and controlled-evidence signal. It is not an IQ test, diagnosis, or proof that readiness routing improves far transfer.

Relational working memory and relational n-back

Generic working-memory training reliably improves trained and structurally similar tasks, but high-quality reviews do not support broad far-transfer claims from ordinary working-memory drill alone (Melby-Lervåg et al., 2016). Trident G therefore places more weight on temporary bindings and relations. Oberauer found that increasing memory load particularly impaired memory for bindings, supporting working memory as a limited workspace for maintaining relational structure (Oberauer, 2019).

Relational integration is also closely associated with fluid reasoning. It predicts reasoning beyond other working-memory tasks, and a latent relation-processing factor has been reported as statistically equivalent to a fluid-reasoning factor in one study (Chuderski, 2014), (Jastrzębski et al., 2020). A recent randomised study trained numerical relations in 1-back and 2-back formats and found changes in frontoparietal EEG microstate dynamics associated with fluid intelligence (Wang et al., 2025).

Meaning for Trident G: relational n-back targets a more reasoning-relevant operation than item repetition because the remembered unit is a relation. Evidence is promising, but it does not establish that relational n-back reliably raises IQ or produces broad far transfer.

Relational reasoning, abstraction, and strategy

Reasoning transfer is more plausible when learners practise identifying invariant relations and explicit strategies rather than repeating one surface form. Reviews of inductive-reasoning training support comparison, rule formation, and testing regularities across examples (Klauer & Phye, 2008). In a smaller randomised study with children, strategy-based working-memory training transferred to a novel problem-solving task, suggesting that portable strategies can matter beyond raw drill (Chan et al., 2019).

Meaning for Trident G: Reasoning Gym’s focus on relation matching, constraints, must-follow inference, and strategy reflection is research-aligned. Transfer is still expected to depend on task design, cueing, and the distance between training and outcome tasks.

Horizontal variation and invariant recovery

Perceptual learning is often highly specific, but transfer can improve when training helps learners recover a higher-level rule or when practice is distributed across time and conditions (Zhang et al., 2010), (Larcombe et al., 2017). This supports changing wrappers while preserving the underlying relation, then checking whether performance recovers after the surface shift.

The entropy–mutual-information principle also draws formal inspiration from Zhang and Tang’s model of neural-network learning. Their analysis describes learning as a balance between maximum-entropy exploration and mutual-information constraint with the task objective (Zhang & Tang, 2025). For Trident G, the analogy is that variation should be broad enough to prevent brittle surface memorisation but constrained enough to preserve the target invariant.

Meaning for Trident G: wrapper swaps are a plausible transfer design and validation method. Zhang and Tang provide a formal analogy from machine learning, not evidence that the same mechanism causes human cognitive transfer.

Vertical progression and real-world deployment

Vertical progression links readiness and attention control to working-memory binding, reasoning strategies, and action. Evidence for this exact chain is not yet direct, but strategy-enriched working-memory interventions are more promising than drill alone, and implementation-intention research shows that cue-linked “if–then” plans can improve goal attainment (Chan et al., 2019), (Gollwitzer & Sheeran, 2006).

Meaning for Trident G: prompts and missions are intended to connect trained control to use outside the exercise. No public IQ Pro mission-success report is currently presented; real-world outcome evidence will be added only when suitable reports exist.

Spacing, consolidation, and delayed re-checks

Distributed practice is a robust route to more durable learning, while consolidation conditions can influence whether perceptual and motor learning generalises beyond the original task (Gerbier & Toppino, 2015), (Censor, 2013). A delayed check therefore carries more information than an immediate post-practice score because it asks whether the trained structure remains accessible after time and changed conditions.

Meaning for Trident G: short sessions, spacing, and later re-checks are evidence-aligned design choices. They improve the quality of a transfer test; they do not guarantee that transfer will occur.