Pith. sign in

Title resolution pending

2 Pith papers cite this work. Polarity classification is still indexing.

2 Pith papers citing it

fields

cs.CL 1 cs.LG 1

years

2026 2

representative citing papers

Prompt Compression via Activation Aggregation

cs.CL · 2026-07-09 · conditional · novelty 6.0

A learned weighted sum of intermediate-layer activations compresses an instruction prompt into a single patch vector that, injected at an early layer, recovers task accuracy within ~2% of the full prompt.

citing papers explorer

Showing 2 of 2 citing papers.

  • From Circuit Evidence to Mechanistic Theory: An Inductive Logic Approach cs.LG · 2026-05-20 · unverdicted · none · ref 63

    Introduces Causal Functional Signatures grounded in causal evidence and ILP-learned architectural signatures to enable explicit, comparable, and portable mechanistic claims across model scales.

  • Prompt Compression via Activation Aggregation cs.CL · 2026-07-09 · conditional · none · ref 8

    A learned weighted sum of intermediate-layer activations compresses an instruction prompt into a single patch vector that, injected at an early layer, recovers task accuracy within ~2% of the full prompt.