Differentiable Weightless Controllers: Learning Logic Circuits for Continuous Control
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866915646992809984 |
|---|---|
| author | Kresse, Fabian Lampert, Christoph H. |
| author_facet | Kresse, Fabian Lampert, Christoph H. |
| contents | We investigate whether continuous-control policies can be represented and learned as discrete logic circuits instead of continuous neural networks. We introduce Differentiable Weightless Controllers (DWCs), a symbolic-differentiable architecture that maps real-valued observations to actions using thermometer-encoded inputs, sparsely connected boolean lookup-table layers, and lightweight action heads. DWCs can be trained end-to-end by gradient-based techniques, yet compile directly into FPGA-compatible circuits with few- or even single-clock-cycle latency and nanojoule-level energy cost per action. Across five MuJoCo benchmarks, including high-dimensional Humanoid, DWCs achieve returns competitive with weight-based policies (full precision or quantized neural networks), matching performance on four tasks and isolating network capacity as the key limiting factor on HalfCheetah. Furthermore, DWCs exhibit structurally sparse and interpretable connectivity patterns, enabling a direct inspection of which input thresholds influence control decisions. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_01467 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Differentiable Weightless Controllers: Learning Logic Circuits for Continuous Control Kresse, Fabian Lampert, Christoph H. Machine Learning Hardware Architecture Symbolic Computation We investigate whether continuous-control policies can be represented and learned as discrete logic circuits instead of continuous neural networks. We introduce Differentiable Weightless Controllers (DWCs), a symbolic-differentiable architecture that maps real-valued observations to actions using thermometer-encoded inputs, sparsely connected boolean lookup-table layers, and lightweight action heads. DWCs can be trained end-to-end by gradient-based techniques, yet compile directly into FPGA-compatible circuits with few- or even single-clock-cycle latency and nanojoule-level energy cost per action. Across five MuJoCo benchmarks, including high-dimensional Humanoid, DWCs achieve returns competitive with weight-based policies (full precision or quantized neural networks), matching performance on four tasks and isolating network capacity as the key limiting factor on HalfCheetah. Furthermore, DWCs exhibit structurally sparse and interpretable connectivity patterns, enabling a direct inspection of which input thresholds influence control decisions. |
| title | Differentiable Weightless Controllers: Learning Logic Circuits for Continuous Control |
| topic | Machine Learning Hardware Architecture Symbolic Computation |
| url | https://arxiv.org/abs/2512.01467 |