Attach a certificate to a cached transformer KV state or recurrent latent state and refresh it only while its predicted certificate remains inside a latency-contracted admissible region. The controller uses a bound on certificate drift to guarantee that the state will remain admissible throughout the next sampling, communication, and execution delay, reducing unnecessary recomputation while exposing a measurable refresh boundary.
Construct overlapping attention windows from a block design instead of using one dense sequence-by-sequence attention matrix. Every token appears in exactly $r$ windows and every token pair co-occurs in exactly $\lambda$ windows, giving uniform coverage and avoiding the uneven connectivity of arbitrary sparse masks.
Use a low-rank Tucker reconstruction as a structured backbone and quantize only its residual after an orthogonal rotation. The rotation preserves residual energy but redistributes it across coordinates, reducing dynamic-range imbalance and making 2- or 4-bit uniform quantization less damaging than direct quantization of the original KV tensor.
Paper: A JoLT for the KV cache: Near-lossless KV cache compression via joint Lagrangian allocation of Tucker ranks and a rotated residual for llmsarXiv:2607.12550
Construct a sparse attention layer by sampling backward token histories as a continuous-time branching process rather than allowing every query to attend to every key. Each active ancestor either dies or branches into a bounded number of candidate ancestors, with branching probability controlled by a small parameter. The branch-out penalty predicts exponentially small probability of long, highly branching histories, providing a direct knob for receptive-field size and attention FLOPs.
Use the sharp exponential tail bound to set a local clipping threshold from a desired exceedance probability. Instead of globally clipping activations at a fixed value or percentile, clip each local window at its minimum plus B log(e/delta), where delta is the tolerated fraction of clipped entries.
Sample one batching aggressiveness parameter once per serving epoch and use it for all busy periods, instead of independently randomizing every batch. This globally coupled randomness preserves nested batching behavior while preventing a fixed traffic pattern from repeatedly aligning with a deterministic timeout.
Use local split-fusion rewrites as a structured alternative to globally recomputing token clusters. A model proposes a small number of neighboring tree edits per input, accepts only valid edits that reduce a learned energy, and retains the previous hierarchy across layers or decoding steps.
Use the one-edit Hamiltonian walk as an explicit state machine for counters in autoregressive models or world models. Instead of regenerating and re-embedding an entire numeric string after every increment, update only the digit that changes, or append the single leading digit at a block transition.