pith. sign in

International Conference on Machine Learning , pages=

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

Bug or Feature$^2$: Weight Drift, Activation Sparsity and Spikes

cs.LG · 2026-05-17 · accept · novelty 7.0 · 2 refs

The paper proves negative weight drift at initialization under MSE or cross-entropy with asymmetric activations, links it to up to 90% sparsity in GPT-nano, maps the sparsity-accuracy cliff across 79 configurations, and shows clipped ReLU² and GELU² improve validation loss.

citing papers explorer

Showing 2 of 2 citing papers.

  • Bug or Feature$^2$: Weight Drift, Activation Sparsity and Spikes cs.LG · 2026-05-17 · accept · none · ref 30 · 2 links

    The paper proves negative weight drift at initialization under MSE or cross-entropy with asymmetric activations, links it to up to 90% sparsity in GPT-nano, maps the sparsity-accuracy cliff across 79 configurations, and shows clipped ReLU² and GELU² improve validation loss.

  • Representation-Guided Parameter-Efficient LLM Unlearning cs.CL · 2026-04-19 · unverdicted · none · ref 134

    REGLU guides LoRA-based unlearning via representation subspaces and orthogonal regularization to outperform prior methods on forget-retain trade-off in LLM benchmarks.