TopK-Guided: Adaptive, Budget-Aware Activation Sparsity for Efficient LLM Inference

arXiv:2610.01763v1 Announce Type: new Abstract: Activation sparsity speeds up large language model (LLM) inference by setting unimportant activations to zero so that the corresponding computations can be skipped. Existing training-free methods, however, make different trade-offs: threshold-based methods such as TEAL adapt the sparsity level to each token but do not tightly control the realised sparsity, while TopK-based methods such as WINA enforce a fixed sparsity level but use the same sparsity budget for every token. Both also apply the same budget across transformer blocks, despite large differences in block sensitivity. We introduce TopK-Guided, a training-free method that addresses both limitations by combining bounded token-level sparsity adaptation with sensitivity-aware block-level budget allocation. Across Llama-2 and Llama-3 models, TopK-Guided consistently improves perplexity and downstream accuracy over TEAL and WINA while preserving essentially the same sparsitydependent projection compute as WINA, with the largest gains at high sparsity. Ablations show that both components provide complementary improvements.

Related

• Minimizing Targeted Activations: Input-Only Suppression of Evaluation-Awareness Latents in Large Language Models
• Near-Oracle KV Selection via Pre-hoc Sparsity for Long-Context Inference
• DistillCache: KL-Guided Adaptive KV-Cache Eviction for Memory-Efficient LLM Inference
• InfoQuant: Shaping Activation Distributions for Low-Bit LLM Quantization

Source: arXiv cs.AI | 2026-10-02