ArcGate: Adaptive Arctangent Gated Activation
Avik Bhattacharya, Siddhant Dnyanesh Gole, Subhasis Chaudhuri, Alejandro C. Frery, Biplab Banerjee · 2026
Read ArcGate: Adaptive Arctangent Gated Activation by Avik Bhattacharya, Siddhant Dnyanesh Gole, Subhasis Chaudhuri, Alejandro C. Frery, Biplab Banerjee (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-14518.txt.
Description
Bhattacharya and colleagues propose ArcGate, an Adaptive Arctangent Gated Activation making shapes through a three-stage non-linear transformation. Unlike fixed-shape activations such as ReLU, GELU or SiLU, ArcGate uses seven learnable weights per layer. On three remote sensing benchmarks, ArcGate beats standard rivals, reaching 99.67 percent overall accuracy on PatternNet. ArcGate also holds up better under moderate Gaussian noise, keeping a 26.65 percent performance lead over ReLU. Learned weights show gating strength rising in deeper layers to enhance signal flow.
Catalog record
| CREATOR | Avik Bhattacharya, Siddhant Dnyanesh Gole, Subhasis Chaudhuri, Alejandro C. Frery, Biplab Banerjee |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 22098 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.14518