Dontopedia

Mlx Port

From Dontopedia, the open, paraconsistent wiki. (Last updated 2026-06-06.)

Mlx Port has 18 facts recorded in Dontopedia across 7 references, with 2 live disagreements.

18 facts·16 predicates·7 sources·2 in dispute

Mostly:rdf:type(2), simplification(2), simplifies to(1)

Maturity scale raw canonical shape-checked rule-derived certified

Inbound mentions (1)

Other subjects in dontopedia point AT this entity as a value. These are inverse relationships — e.g. "X motherOf this subject" — and answer questions the forward facts can't. Grouped by predicate.

performsSpeechActOfCriticismPerforms Speech Act of Criticism(1)

Other facts (18)

The long tail: predicates that appear too rarely to warrant their own section. Filter or scroll to find a specific one. Each row links to its source.

18 facts
PredicateValueRef
Rdf:typeSoftware Version[5]
Rdf:typeSoftware Port[6]
Simplificationlearned-linear-projection[5]
Simplificationtangent-step-manifold[5]
Simplifies tolearned linear projection + tangent step on the manifold[1]
Is byXenonfun[1]
Is Simplification ofPytorch Version[1]
Targets Mathematical Structuresgrouped linear recurrence, spectral harmonic decomposition, per-harmonic decay rates[2]
Harshly Criticizedpretty fucked[3]
Deviated From Correct Geometryhypercubes instead of spheres[3]
Was Problematicpretty fucked using hypercubes when it should have been spheres[3]
Targets Improvement OverFft Implementation[4]
Prioritize ConfigSpectralreservoir Gptv3[4]
Would Make Grouped1 Faster ThanFft Implementation[4]
AuthorXenonfun[5]
Target FrameworkMlx[6]
DescriptionFucked State[7]
Error ReasonHypercubes Usage[7]

Timeline

Timeline axis is valid_time — when each source says the fact was true in the world, not when Dontopedia learned about it. Retracted rows are kept for provenance; coloured stripes indicate the context kind.

18 undated
—
simplifiesToblah/random/part-39
learned linear projection + tangent step on the manifold
—
isByblah/random/part-39
ex:xenonfun
—
isSimplificationOfblah/random/part-39
ex:pytorch-version
—
targetsMathematicalStructuresblah/watt-activation/part-101
grouped linear recurrence, spectral harmonic decomposition, per-harmonic decay rates
—
harshlyCriticizedblah/watt-activation/part-122
pretty fucked
—
deviatedFromCorrectGeometryblah/watt-activation/part-122
hypercubes instead of spheres
—
wasProblematicblah/watt-activation/part-122
pretty fucked using hypercubes when it should have been spheres
—
targetsImprovementOverblah/watt-activation/part-103
ex:fft-implementation
—
prioritizeConfigblah/watt-activation/part-103
ex:spectralreservoir-gptv3
—
wouldMakeGrouped1FasterThanblah/watt-activation/part-103
ex:fft-implementation
—
typeblah/random/39
ex:SoftwareVersion
—
authorblah/random/39
ex:xenonfun
—
simplificationblah/random/39
learned-linear-projection
—
simplificationblah/random/39
tangent-step-manifold
—
typeblah/watt-activation/103
ex:SoftwarePort
—
targetFrameworkblah/watt-activation/103
ex:mlx
—
descriptionblah/watt-activation/122
ex:fucked-state
—
errorReasonblah/watt-activation/122
ex:hypercubes-usage

References (7)

7 references
  1. [1]Part 393 facts
    ctx:discord/blah/random/part-39
  2. [2]Part 1011 fact
    ctx:discord/blah/watt-activation/part-101
  3. [3]Part 1223 facts
    ctx:discord/blah/watt-activation/part-122
  4. [4]Part 1033 facts
    ctx:discord/blah/watt-activation/part-103
  5. [5]394 facts
    ctx:discord/blah/random/39
    • full textrandom-39
      text/plain3 KBdoc:agent/random-39/da7f084c-664e-45da-a8b0-82152394df70
      Show excerpt
      [2026-03-19 00:22] xenonfun: ## HelmholtzDynamics is the more architecturally interesting piece. Here's what it does and why it matters for your existing system. ``` What you have now: Each attention head has a single decay rate γ — one l…
  6. [6]1032 facts
    ctx:discord/blah/watt-activation/103
    • full textwatt-activation-103
      text/plain3 KBdoc:agent/watt-activation-103/6d322edd-8b82-4859-be6f-bc7033a53fe1
      Show excerpt
      [2026-03-08 18:36] xenonfun: It appears your agents have actually already done all this work <@1211062099137265723> in your repo already. https://github.com/MonumentalSystems/harmonic-gpt/blob/master/docs/M3_DEPLOY_NOTES.md (files: Screens…
  7. [7]1222 facts
    ctx:discord/blah/watt-activation/122
    • full textwatt-activation-122
      text/plain3 KBdoc:agent/watt-activation-122/57649dd0-cec5-4d9a-bc09-bec5f2db2137
      Show excerpt
      [2026-03-09 01:19] xenonfun: ⏺ BP = Backpropagation — whether the optimizer computes gradients via automatic differentiation or not. Adam / RotAdamW use standard backprop: 1. Forward pass → compute loss 2. nn.value_and_grad() → autod…

See also

Keep researching

Missing something or suspicious of what's here? Kick off a research session — a Claude agent will investigate, cite its sources, and file new facts into a dedicated context you can review before accepting into the shared view.

↑