#latent — Public Fediverse posts on home.social

deepseek @[email protected] · 2026-03-09 · 12:45 UTC

DeepSeek-V3 Model: Theory, Config, and Rotary Positional Embeddings Table of Contents DeepSeek-V3 Model: Theory, Config, and Rotary Positional Embeddings Introduction to the DeepSeek-V3 Model The F...

#DeepSeek-V3 #KV #Cache #MultiHead #Latent #Attention #RoPE #Tutorial #deepseekv3 #kv #cache

Origin | Interest | Match

#deepseekv3 #kv #cache #multihead #latent #attention

deepseek @[email protected] · 2025-10-13 · 13:00 UTC

KV Cache Optimization via Multi-Head Latent Attention Table of Contents KV Cache Optimization via Multi-Head Latent Attention Recap of KV Cache The Need for KV Cache Optimization Multi-Head Latent ...

#KV #Cache #LLM #Inference #LLMs #Multi-Head #Latent #Attention #MultiHead #Attention #Tutorial

Origin | Interest | Match

#kv #cache #llm #inference #llms #multihead

deepseek @[email protected] · 2025-10-13 · 13:00 UTC

KV Cache Optimization via Multi-Head Latent Attention Table of Contents KV Cache Optimization via Multi-Head Latent Attention Recap of KV Cache The Need for KV Cache Optimization Multi-Head Latent ...

#KV #Cache #LLM #Inference #LLMs #Multi-Head #Latent #Attention #MultiHead #Attention #Tutorial

Origin | Interest | Match

#kv #cache #llm #inference #llms #multihead

Fabrizio Musacchio @[email protected] · 2025-08-04 · 05:02 UTC

In our recent #JournalClub, I presented Genkin et al. (2025), who decode #DecisionMaking in the #PremotorCortex of #macaques as low-dimensional #latent #dynamics shared across #NeuralPopulations. Their generative model links tuning curves, spike-time variability, and stimulus-dependent potential landscapes to a common internal decision variable. I summarized and discussed their findings in this blog post:

📝https://doi.org/10.1038/s41586-025-09199-1
🌍https://www.fabriziomusacchio.com/blog/2025-08-01-decoding_decision_making_in_premotor_cortex/

#CompNeuro #Neuroscience #Cortex

#journalclub #decisionmaking #premotorcortex #macaques #latent #dynamics

Fabrizio Musacchio @[email protected] · 2025-07-24 · 08:16 UTC

And as an added bonus: Dileep George, one of the authors of the paper, just shared a #JupyterNotebook demo 🐍📔. You can explore the #CSCG model, visualize #PlaceFields, and inspect the learned #latent #graphs 📈 Just try it out, it's great fun 👌

🌍 https://colab.research.google.com/drive/1kgjuoz_Noo7uV87StSbW7T8-IBQmPOLE?usp=sharing#scrollTo=7jnInNH7RUAX

#Hippocampus #CognitiveMaps #CSCG #Neuroscience #ComputationalModeling #CompNeuro

#jupyternotebook #cscg #placefields #latent #graphs #hippocampus

Fabrizio Musacchio @[email protected] · 2025-07-24 · 05:15 UTC

This paper by Raju et al. proposes a unified model – “clone‑structured causal #graphs” (#CSCG) – for #hippocampal #SpatialCoding. It suggests that #SpatialMaps arise from #learning #latent higher‑order sequences rather than representing #EuclideanSpace directly. The model elegantly explains phenomena like #PlaceFields, #SplitterCells, #contextual #remapping, and predicts when #PlaceFieldMapping may mislead.

🌍 https://www.science.org/doi/10.1126/sciadv.adm8470

#Hippocampus #CognitiveMaps #SequenceLearning #Neuroscience

#graphs #cscg #hippocampal #spatialcoding #spatialmaps #learning

Fabrizio Musacchio @[email protected] · 2025-07-24 · 05:15 UTC

This paper by Raju et al. proposes a unified model – “clone‑structured causal #graphs” (#CSCG) – for #hippocampal #SpatialCoding. It suggests that #SpatialMaps arise from #learning #latent higher‑order sequences rather than representing #EuclideanSpace directly. The model elegantly explains phenomena like #PlaceFields, #SplitterCells, #contextual #remapping, and predicts when #PlaceFieldMapping may mislead.

🌍 https://www.science.org/doi/10.1126/sciadv.adm8470

#Hippocampus #CognitiveMaps #SequenceLearning #Neuroscience

#graphs #cscg #hippocampal #spatialcoding #spatialmaps #learning

Fabrizio Musacchio @[email protected] · 2025-07-24 · 05:15 UTC

This paper by Raju et al. proposes a unified model – “clone‑structured causal #graphs” (#CSCG) – for #hippocampal #SpatialCoding. It suggests that #SpatialMaps arise from #learning #latent higher‑order sequences rather than representing #EuclideanSpace directly. The model elegantly explains phenomena like #PlaceFields, #SplitterCells, #contextual #remapping, and predicts when #PlaceFieldMapping may mislead.

🌍 https://www.science.org/doi/10.1126/sciadv.adm8470

#Hippocampus #CognitiveMaps #SequenceLearning #Neuroscience

#graphs #cscg #hippocampal #spatialcoding #spatialmaps #learning

Fabrizio Musacchio @[email protected] · 2025-07-24 · 05:15 UTC

This paper by Raju et al. proposes a unified model – “clone‑structured causal #graphs” (#CSCG) – for #hippocampal #SpatialCoding. It suggests that #SpatialMaps arise from #learning #latent higher‑order sequences rather than representing #EuclideanSpace directly. The model elegantly explains phenomena like #PlaceFields, #SplitterCells, #contextual #remapping, and predicts when #PlaceFieldMapping may mislead.

🌍 https://www.science.org/doi/10.1126/sciadv.adm8470

#Hippocampus #CognitiveMaps #SequenceLearning #Neuroscience

#neuroscience #sequencelearning #cognitivemaps #hippocampus #placefieldmapping #remapping

Fabrizio Musacchio @[email protected] · 2025-07-24 · 05:15 UTC

This paper by Raju et al. proposes a unified model – “clone‑structured causal #graphs” (#CSCG) – for #hippocampal #SpatialCoding. It suggests that #SpatialMaps arise from #learning #latent higher‑order sequences rather than representing #EuclideanSpace directly. The model elegantly explains phenomena like #PlaceFields, #SplitterCells, #contextual #remapping, and predicts when #PlaceFieldMapping may mislead.

🌍 https://www.science.org/doi/10.1126/sciadv.adm8470

#Hippocampus #CognitiveMaps #SequenceLearning #Neuroscience

#graphs #cscg #hippocampal #spatialcoding #spatialmaps #learning