Skip to content

Latest commit

 

History

History
39 lines (29 loc) · 4.09 KB

File metadata and controls

39 lines (29 loc) · 4.09 KB

Model Zoo

All models integrated into SpikingTSF are listed below. Model implementations are adapted from the source repositories indicated. This library does not claim ownership of any original architecture.

When using a specific model, please cite both SpikingTSF and the corresponding source paper.


Model Type Spike? Source Paper Source Repository File in Repo Status
SpikF-GO Spiking Fourier Graph Operator (freq. domain, optional spike-domain CPG-PE) Bakhshaliyev & Landwehr, accepted to ECML PKDD 2026 (our own paper — link to be added on publication) jafarbakhshaliyev/SpikF-GO models/SpikF_GO.py Runnable, ETTh1+ETTh2
SpikF Transformer (freq. domain) Wu et al., ICML 2025 WWJ-creator/SpikF models/SpikF.py Runnable, ETTh1+ETTh2
Spikformer Transformer (spike-driven SA) SeqSNN, NeurIPS 2025 microsoft/SeqSNN models/Spikformer.py Runnable, ETTh1
Spikingformer Transformer (pre-LIF) SeqSNN, NeurIPS 2025 microsoft/SeqSNN models/Spikingformer.py Runnable, ETTh1+ETTh2 verified
QKFormer Transformer (token-level Q/K attn) SeqSNN, NeurIPS 2025 microsoft/SeqSNN models/QKFormer.py Runnable, results pending
TSGRU GRU (two-compartment TS-LIF) TS-LIF, ICLR 2025 kkking-kk/TS-LIF models/TSGRU.py Runnable, ETTh1+ETTh2 verified
TSTCN TCN (two-compartment TS-LIF) TS-LIF, ICLR 2025 kkking-kk/TS-LIF models/TSTCN.py Runnable, ETTh1 verified
TSFormer Transformer (two-compartment TS-LIF) TS-LIF, ICLR 2025 kkking-kk/TS-LIF models/TSFormer.py Runnable, ETTh1 verified
iSpikformer Inverted Spiking Transformer SeqSNN, ICML 2024 microsoft/SeqSNN models/iSpikformer.py Runnable, ETTh1 verified
SpikeRNN Spiking Recurrent Network SeqSNN, ICML 2024 microsoft/SeqSNN models/SpikeRNN.py Runnable, ETTh1+ETTh2 verified
SpikTCN Spiking Temporal Convolutional Network SeqSNN, ICML 2024 microsoft/SeqSNN models/SpikTCN.py Runnable, ETTh1 partial
SpikGRU Spiking Gated Recurrent Unit SeqSNN, ICML 2024 microsoft/SeqSNN models/SpikGRU.py Runnable, ETTh1+ETTh2 verified
iTransformer Inverted Transformer (ANN) Liu et al., ICLR 2024 thuml/iTransformer models/ITransformer.py Runnable, results pending
DLinear Decomposition Linear (ANN) Zeng et al., AAAI 2023 thuml/Time-Series-Library models/DLinear.py Runnable, results pending

Status Definitions

  • Runnable: Code is integrated and can be executed with the provided scripts.
  • Verified: Results have been produced and logged under Output/. Numbers are in RESULTS.md.
  • Partial: Some horizons or datasets have been evaluated but not all.
  • Pending: Code is present but results have not yet been computed.

Adding a New Model

See CONTRIBUTING.md and docs/adding_a_model.md.