Every idea extracted from recent arXiv mathematics papers — verified and unverified. Click an idea to open its full card; badges show the empirical verdict.
Add an auxiliary loss that makes selected representation coordinates insensitive to all subsets of fewer than d variables while retaining a d-way parity statistic. The objective discourages the network from solving a task through pairwise shortcuts and explicitly rewards a controlled high-order interaction.
Adapt the slope of each spiking neuron's surrogate derivative using the normalized entropy of its block's attention distribution. High centered entropy uncertainty increases the slope, while low uncertainty decreases it, and a dead zone holds the default slope fixed for ordinary fluctuations. The adaptation exists only in backpropagation, so the forward spike function, parameter count, and inference cost remain unchanged.
Add a learnable orthogonal rotation to a hidden representation and train it to make every channel projection have a small ψ2/L2 ratio. Unlike variance normalization, this explicitly suppresses directions with unusually heavy empirical tails while preserving the total quadratic energy of the representation.
Apply a low-degree polynomial feature lift to normalized hidden representations and penalize degeneracy of the covariance in that lifted space. This can detect collapse in nonlinear combinations of features even when the raw hidden covariance appears healthy.
Replace a collection of dense task-specific linear layers with a common sparse structural matrix and task-specific edge strengths. All tasks share the same learned connectivity pattern, but retain independent values on active connections, allowing parameter sharing without forcing identical interactions.
Replace arbitrary learned thresholds in a binary MoE or hierarchical latent router with a threshold at the batch mean of a learned scalar projection. Add a penalty when the entropy of either routed subgroup falls too far below the parent entropy, using the paper's sharp constant as the target. This discourages routing branches from becoming nearly deterministic or semantically impoverished while retaining a simple, cheap gating operation.
Treat the maximum dependency distance faithfully modeled by a finite neural architecture as an emergent correlation length, and estimate how it grows with depth, state size, or attention span. Fit the exponent \(\kappa\) and use it as an architecture-selection signal: a model with larger \(\kappa\) should acquire long-range competence more efficiently at equal parameter or FLOP budget.
Normalize attention or router logits and control their upper tail using the paper's sharper Gaussian-maximum exponent rather than a correlation-blind sub-Gaussian bound. Use the resulting threshold to add a soft penalty or adaptive temperature whenever the observed maximum exceeds the calibrated level, reducing rare one-token or one-expert domination.
Replace softmax attention or dense MoE routing with a normalized Rayleigh–Jeans distribution over tokens or experts. If an item's energy is close to the chemical potential, its probability becomes disproportionately large, creating controllable low-energy condensation instead of the exponentially smooth allocation produced by softmax.
Replace a neural layer's dense parameters by a real parameter tensor and a binary activity mask, then update the mask using importance scores measured across multiple perturbation intervals. Gates that repeatedly occur in non-dominated accuracy–cost configurations receive larger scores and are more likely to be retained or activated, avoiding uniformly random sparse rewiring.
Calibrate the maximum attention logit in each head against the log-correlated extreme-value law instead of applying fixed clipping or a fixed max-norm penalty. Penalize only maxima that exceed the predicted log N minus three-quarter log log N baseline by an unusually large order-one fluctuation, allowing ordinary sharp attention while suppressing rare pathological spikes.