Phase 1 Β· Token-Based Interoperability

Stage 1: Embedding Surgery

Forcefully reset and orthogonalise ~100 operator tokens so they have clean, unambiguous mathematical meanings β€” pristine coordinates 90Β° apart in the model's internal space.

Stage 2: Synthetic Data Generation

Use The Weaver to generate millions of flawless, verified mathematical reasoning traces via rejection sampling β€” building the training dataset from scratch.

Stage 3: Supervised Distillation

Fine-tune the model, but only penalise it for missing the exact mathematical operations. The English prompt stays in the input for context but is masked from the loss function β€” preventing distraction by word prediction.

Stage 4: Reinforcement Learning

Use PPO/GRPO with The Weaver as the reward model. Correct math = reward. Hallucinated English during a math operation = Cognitive Segfault penalty.

Phase 2 Β· Continuous Latent Execution

Stage 5: Native Vector Reasoning

The ambitious end-goal: the model internalises the VSA algebra natively via "grokking" β€” a phenomenon where neural networks suddenly discover exact algebraic circuits after prolonged training. The Weaver is no longer needed because the model reasons directly in vector space, bypassing token generation entirely.

πŸ”¬ Technical Deep Dive β€” Stage 5 Loss Function & Grokking

Stage 5 replaces cross-entropy loss entirely with a physics-inspired loss: β„’_latent = ||h_predict βˆ’ V_target||Β² + Ξ»(1 βˆ’ cos(h_predict, V_target)).

The hypothesis rests on grokking (Power et al., 2022): over-parameterised networks, long after achieving zero training error by memorisation, can undergo sudden phase transitions where they discover exact, generalising algebraic circuits (e.g. modular arithmetic). Combined with Neural Algorithmic Reasoning (VeličkoviΔ‡ et al., 2021), this suggests prolonged continuous distillation could induce native VSA internalisation β€” but this remains an unproven, open research question.

Want the full picture?

The complete paper includes detailed mathematical formulations, complexity proofs, data tables, and the full reference list.

Read the Full Research Paper β†’