# Community

**URL:** https://discourse.numenta.org/c/community/49.md

[Latest](https://discourse.numenta.org/latest.md) · [Categories](https://discourse.numenta.org/categories.md) · [Tags](https://discourse.numenta.org/tags.md)

---

## [About the Community category](https://discourse.numenta.org/t/about-the-community-category/6198)

<div class="topic-metadata">

**Author:** [@rhyolight](https://discourse.numenta.org/u/rhyolight)\
**Replies:** 0

</div>

A place for people interested in HTM and related theories of intelligence to chat about anything they like. Meet new people, start conversations, have fun.

---

## [Broad "If" statements in neural networks from context](https://discourse.numenta.org/t/broad-if-statements-in-neural-networks-from-context/12625)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 3\
**Last updated:** [October 2, 2026, 4:30pm UTC](https://discourse.numenta.org/t/broad-if-statements-in-neural-networks-from-context/12625 "2026-10-02T16:30:32Z")

</div>

If you accept (or are willing to hypothesize) that a neural network is a sequence of context selected linear mappings, then applying SVD to the matrices is very telling: https://archive.org/details/the-broad-if-statemen…

---

## [Numenta Website?](https://discourse.numenta.org/t/numenta-website/12628)

<div class="topic-metadata">

**Author:** [@flajann2](https://discourse.numenta.org/u/flajann2)\
**Replies:** 1\
**Last updated:** [September 28, 2026, 1:53pm UTC](https://discourse.numenta.org/t/numenta-website/12628 "2026-09-28T13:53:14Z")

</div>

It would appear that something is up with the numenta.com website is gone. Domain name issue that happened 4 days ago. Did I miss something?

---

## [Neural Topographic Maps & CCSLM](https://discourse.numenta.org/t/neural-topographic-maps-ccslm/12612)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 9\
**Last updated:** [September 24, 2026, 11:00pm UTC](https://discourse.numenta.org/t/neural-topographic-maps-ccslm/12612 "2026-09-24T23:00:14Z")

</div>

I recalled that there are sensory topographic maps in the brain. Why would the brain do that, unless it wanted to apply a sequence of not too drastic geometric mapping to the data? Perhaps along the lines of linear mappi…

---

## [Linear Mappings have a Memory](https://discourse.numenta.org/t/linear-mappings-have-a-memory/12585)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 9\
**Last updated:** [September 11, 2026, 3:04pm UTC](https://discourse.numenta.org/t/linear-mappings-have-a-memory/12585 "2026-09-11T15:04:39Z")

</div>

Note about the memory effects seen when you use linear mappings as associative memory. Prior memories can hang around for a long time, including random memories programmed in by random initialization. That’s because st…

---

## [Self-improving intelligence](https://discourse.numenta.org/t/self-improving-intelligence/5613)

<div class="topic-metadata">

**Author:** [@Paul\_Lamb](https://discourse.numenta.org/u/Paul_Lamb)\
**Replies:** 46\
**Last updated:** [September 9, 2026, 9:28pm UTC](https://discourse.numenta.org/t/self-improving-intelligence/5613 "2026-09-09T21:28:32Z")

</div>

I am attempting to create a low-level “organism” of sorts, which has all the basic components needed to control a body, form models of itself, of its world, etc. and learn to satisfy needs. This seed entity will be the…

---

## [Neuron Populations: Mathematical Structure of Data and Operations](https://discourse.numenta.org/t/neuron-populations-mathematical-structure-of-data-and-operations/12538)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 12\
**Last updated:** [September 9, 2026, 1:09pm UTC](https://discourse.numenta.org/t/neuron-populations-mathematical-structure-of-data-and-operations/12538 "2026-09-09T13:09:28Z")

</div>

This paper on the biological brain seems to fit in with the concatenated context selected linear mapping viewpoint of artificial neural networks. https://arxiv.org/html/2309.02332v3 See what you think anyway.

---

## [Explicit Context Variables for Context-Conditioned Linear Systems](https://discourse.numenta.org/t/explicit-context-variables-for-context-conditioned-linear-systems/12511)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 8\
**Last updated:** [August 17, 2026, 1:42pm UTC](https://discourse.numenta.org/t/explicit-context-variables-for-context-conditioned-linear-systems/12511 "2026-08-17T13:42:28Z")

</div>

If you think a context dependent chain of linear mappings is a good approach then you have a lot of opportunity to include context signals often seen in biological systems: https://archive.org/details/technical-note-exp…

---

## [Lessons from CCSLM for Understanding Animal Neural Networks](https://discourse.numenta.org/t/lessons-from-ccslm-for-understanding-animal-neural-networks/12529)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 0\
**Last updated:** [August 10, 2026, 3:05am UTC](https://discourse.numenta.org/t/lessons-from-ccslm-for-understanding-animal-neural-networks/12529 "2026-08-10T03:05:24Z")

</div>

Experiments and discussions around concatenated context-selected linear mapping systems suggest several constraints that may be relevant to biological neural networks. https://archive.org/details/lessons-from-concatenat…

---

## [From Piecewise Linear Networks to Context Selected Linear Mappings](https://discourse.numenta.org/t/from-piecewise-linear-networks-to-context-selected-linear-mappings/12520)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 0\
**Last updated:** [August 1, 2026, 10:07pm UTC](https://discourse.numenta.org/t/from-piecewise-linear-networks-to-context-selected-linear-mappings/12520 "2026-08-01T22:07:45Z")

</div>

It is widely understood that ReLU neural networks are piece-wise linear mappings. I have tried to give an ‘almost no choice’ argument leading to a layer-wise context mechanism: From Piecewise Linear Networks to Context…

---

## [Context Dependent Composition Of Linear Operators](https://discourse.numenta.org/t/context-dependent-composition-of-linear-operators/12510)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 0\
**Last updated:** [July 22, 2026, 11:54pm UTC](https://discourse.numenta.org/t/context-dependent-composition-of-linear-operators/12510 "2026-07-22T23:54:09Z")

</div>

I made an effort to express in a comprehensible way some ideas I have about context dependent composition of linear operators: https://archive.org/details/context-dependent-composition-of-linear-operators The funny thi…

---

## [BDH (Baby Dragon Hatchling)](https://discourse.numenta.org/t/bdh-baby-dragon-hatchling/12185)

<div class="topic-metadata">

**Author:** [@MTIzNDU2Nzg5](https://discourse.numenta.org/u/MTIzNDU2Nzg5)\
**Replies:** 4\
**Last updated:** [July 21, 2026, 10:37am UTC](https://discourse.numenta.org/t/bdh-baby-dragon-hatchling/12185 "2026-07-21T10:37:30Z")

</div>

Https://arxiv.org/pdf/2509.26507 This paper on BDH (Baby Dragon Hatchling). It’s basically an attempt to replace the Transformer bottleneck with a scale-free graph of “neuron particles.” The technical gist: No KV Ca…

---

## [Context based learning and the best average linear mapping](https://discourse.numenta.org/t/context-based-learning-and-the-best-average-linear-mapping/12465)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 3\
**Last updated:** [July 16, 2026, 10:34pm UTC](https://discourse.numenta.org/t/context-based-learning-and-the-best-average-linear-mapping/12465 "2026-07-16T22:34:12Z")

</div>

Given a context at each layer that selects parameters for a linear mapping, training produces the best average linear mapping for that context. Then the entire networks is a composition of best average linear mappings. …

---

## [The Triad of Context Driven Learning](https://discourse.numenta.org/t/the-triad-of-context-driven-learning/12458)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 1\
**Last updated:** [July 8, 2026, 12:35pm UTC](https://discourse.numenta.org/t/the-triad-of-context-driven-learning/12458 "2026-07-08T12:35:57Z")

</div>

I wrote up some details of context driven learning in as simple a way as I could: https://archive.org/details/triad-of-context-driven-learning The triad is: Context Conditional computing Learning

---

## [Atlas LSH neural networks](https://discourse.numenta.org/t/atlas-lsh-neural-networks/12457)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 5\
**Last updated:** [July 3, 2026, 12:32am UTC](https://discourse.numenta.org/t/atlas-lsh-neural-networks/12457 "2026-07-03T00:32:16Z")

</div>

I’ve been exploring some ideas about locality sensitive hash based neural networks. https://archive.org/details/atlas-lsh-neural-networks-an-intuitive-overview You can click on “uploaded by” for more information.

---

## [Random Projection Sub-Sampling and Restoration](https://discourse.numenta.org/t/random-projection-sub-sampling-and-restoration/12431)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 2\
**Last updated:** [May 27, 2026, 5:08am UTC](https://discourse.numenta.org/t/random-projection-sub-sampling-and-restoration/12431 "2026-05-27T05:08:19Z")

</div>

One thing I noticed experimenting with associative memory is splitting the data into 2 information orthogonal paths improved the results a lot. Therefore why not split it into multiple orthogonal paths using say random …

---

## [Richer than ReLU](https://discourse.numenta.org/t/richer-than-relu/12364)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 1\
**Last updated:** [May 9, 2026, 10:08pm UTC](https://discourse.numenta.org/t/richer-than-relu/12364 "2026-05-09T22:08:57Z")

</div>

ReLU is great and all that but you are just zeroing channels based on decisions (x\>0?) You could created interactions between the decisions and use those for weight matrix interaction. https://archive.org/details/hd-dh…

---

## [Column selection in ReLU neural networks induces associative memory](https://discourse.numenta.org/t/column-selection-in-relu-neural-networks-induces-associative-memory/12360)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 6\
**Last updated:** [April 27, 2026, 7:28am UTC](https://discourse.numenta.org/t/column-selection-in-relu-neural-networks-induces-associative-memory/12360 "2026-04-27T07:28:25Z")

</div>

I seem to be on some kind of documentation spree. https://archive.org/details/column-selection-in-re-lu-networks-as-hashing These could be helpful: https://archive.org/details/associative-memory-as-a-hashed-linear-rea…

---

## [Data dependent operator selection layers](https://discourse.numenta.org/t/data-dependent-operator-selection-layers/12359)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 0\
**Last updated:** [April 22, 2026, 9:55pm UTC](https://discourse.numenta.org/t/data-dependent-operator-selection-layers/12359 "2026-04-22T21:55:50Z")

</div>

I mentioned a while ago the notion of a very basic neural network being the product of data dependent matrix choices. Eg. A₄B₃B₂A₁x for data dependent choices of A or B. I kinda extended the idea here. https://archive.…

---

## [Two Pathway ReLU neural networks](https://discourse.numenta.org/t/two-pathway-relu-neural-networks/12358)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 0\
**Last updated:** [April 20, 2026, 1:14pm UTC](https://discourse.numenta.org/t/two-pathway-relu-neural-networks/12358 "2026-04-20T13:14:00Z")

</div>

I created this document. https://archive.org/details/2-pathway-re-lu-big-picture It is says a number of things. An ordinary ReLU neural network breaks down the input space into a number of piece-wise linear regions. H…

---

## [WHD Mixing With Two Path ReLU](https://discourse.numenta.org/t/whd-mixing-with-two-path-relu/12356)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 0\
**Last updated:** [April 19, 2026, 1:32am UTC](https://discourse.numenta.org/t/whd-mixing-with-two-path-relu/12356 "2026-04-19T01:32:28Z")

</div>

One reason the fast Walsh Hadamard transform may be useful in neural networks is more efficient (less correlated) parameter usage. Looked at here from the 2 path ReLU perspective: https://archive.org/details/technical-n…

---

## [Decision Matrices, Rank, And Information Flow In Neural Networks](https://discourse.numenta.org/t/decision-matrices-rank-and-information-flow-in-neural-networks/12355)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 0\
**Last updated:** [April 17, 2026, 12:31pm UTC](https://discourse.numenta.org/t/decision-matrices-rank-and-information-flow-in-neural-networks/12355 "2026-04-17T12:31:05Z")

</div>

Decision Matrices, Rank, And Information Flow In Neural Networks: https://archive.org/details/decision-matrices-rank-and-information-flow-in-neural-networks You can click on uploaded by for related documents.

---

## [Fast Walsh Hadamard transform for global mixing in neural networks](https://discourse.numenta.org/t/fast-walsh-hadamard-transform-for-global-mixing-in-neural-networks/12352)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 1\
**Last updated:** [April 11, 2026, 4:25am UTC](https://discourse.numenta.org/t/fast-walsh-hadamard-transform-for-global-mixing-in-neural-networks/12352 "2026-04-11T04:25:05Z")

</div>

I suppose this is getting exhausting https://archive.org/details/fast-walsh-hadamard-transform-in-neural-networks-clean-view The idea is you can make local computations in a neural network layer global with a fast tran…

---

## [Gated Linear Associative Memory](https://discourse.numenta.org/t/gated-linear-associative-memory/12300)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 18\
**Last updated:** [April 9, 2026, 10:54pm UTC](https://discourse.numenta.org/t/gated-linear-associative-memory/12300 "2026-04-09T22:54:21Z")

</div>

I wrote this about a weight matrix being simple linear associative memory and then ReLU functions acting as gates resulting in context based sub-selection of weights in the weight matrix. So then you are really getting …

---

## [The Geometry inside a neural network - artificial or biological](https://discourse.numenta.org/t/the-geometry-inside-a-neural-network-artificial-or-biological/12267)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 4\
**Last updated:** [March 13, 2026, 10:10am UTC](https://discourse.numenta.org/t/the-geometry-inside-a-neural-network-artificial-or-biological/12267 "2026-03-13T10:10:55Z")

</div>

There have been a number of papers about the emergence of geometric form inside neural networks over the past few years: Here is one of the later ones: Reddit - The heart of the internet

---

## [Tiny Recursion Models](https://discourse.numenta.org/t/tiny-recursion-models/12289)

<div class="topic-metadata">

**Author:** [@MTIzNDU2Nzg5](https://discourse.numenta.org/u/MTIzNDU2Nzg5)\
**Replies:** 0\
**Last updated:** [February 13, 2026, 5:27am UTC](https://discourse.numenta.org/t/tiny-recursion-models/12289 "2026-02-13T05:27:29Z")

</div>

---

## [Fused layer neural network](https://discourse.numenta.org/t/fused-layer-neural-network/12202)

<div class="topic-metadata">

**Author:** [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Replies:** 3\
**Last updated:** [February 10, 2026, 1:36pm UTC](https://discourse.numenta.org/t/fused-layer-neural-network/12202 "2026-02-10T13:36:10Z")

</div>

SWNet16 neural network: https://archive.org/details/sw-net-16-b Fuses multiple width 16 CReLU layers into one larger layer using the one-to-all connectivity of a fast transform. Then stacks those layers into a neural n…

---

## [Qualia Disqualified](https://discourse.numenta.org/t/qualia-disqualified/8723)

<div class="topic-metadata">

**Author:** [@EEProf](https://discourse.numenta.org/u/EEProf)\
**Replies:** 52\
**Last updated:** [February 2, 2026, 7:26pm UTC](https://discourse.numenta.org/t/qualia-disqualified/8723 "2026-02-02T19:26:00Z")

</div>

Here you go, chew on this: It’s not hard to see how philosophers have tied themselves into such knots over qualia. They started where anyone with any sense would start: with their strongest and clearest intuitions about…

---

## [🚀 Meet Monty 2026: Online Meetup Invitation Dec 17th @ 9AM PST](https://discourse.numenta.org/t/meet-monty-2026-online-meetup-invitation-dec-17th-9am-pst/12174)

<div class="topic-metadata">

**Author:** [@codeallthethingz](https://discourse.numenta.org/u/codeallthethingz)\
**Replies:** 4\
**Last updated:** [December 19, 2025, 11:59pm UTC](https://discourse.numenta.org/t/meet-monty-2026-online-meetup-invitation-dec-17th-9am-pst/12174 "2025-12-19T23:59:31Z")

</div>

Hi HTM Community :waving\_hand: I’m Will, the community manager over at the Thousand Brains Project. We’d like to invite you all to a meetup next week (which will be recorded and posted on YouTube if you can’t make it) w…

---

## [Tried feeding Grok the forum with a question](https://discourse.numenta.org/t/tried-feeding-grok-the-forum-with-a-question/12113)

<div class="topic-metadata">

**Author:** [@BrainVx](https://discourse.numenta.org/u/BrainVx)\
**Replies:** 2\
**Last updated:** [November 20, 2025, 12:00am UTC](https://discourse.numenta.org/t/tried-feeding-grok-the-forum-with-a-question/12113 "2025-11-20T00:00:33Z")

</div>

Thought this might be of interest and stoke some ideas. Having used Grok for a number of different areas (legal, code, physics, conceptual, etc.) I thought I would see what Grok made of the site and the posts in relatio…

[Next page](https://discourse.numenta.org/c/community/49.md?page=1)