Dontopedia

Experiments

From Dontopedia, the open, paraconsistent wiki. (Last updated 2026-06-11.)

Experiments has 55 facts recorded in Dontopedia across 28 references, with 4 live disagreements.

55 facts·43 predicates·28 sources·4 in dispute

Mostly:include(9), rdf:type(3), ongoing(2)

Maturity scale raw canonical shape-checked rule-derived certified

Inbound mentions (23)

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.

assertsStatusUpdateAsserts Status Update(1)

confirmsStabilityConfirms Stability(1)

considersConsiders(1)

enablesEnables(1)

expressesIntentToMonitorExpresses Intent to Monitor(1)

firesReliablyFires Reliably(1)

hasFeatureHas Feature(1)

identifiedBiggestUseCaseIdentified Biggest Use Case(1)

illustrateResultsIllustrate Results(1)

livesInProjectLives in Project(1)

mentionsMentions(1)

needsToGrowNeeds to Grow(1)

packsMultiplePacks Multiple(1)

perceivesAllShapesAsPerceives All Shapes As(1)

presupposedAsKeyPerformanceIndicatorPresupposed As Key Performance Indicator(1)

providesStatusUpdateProvides Status Update(1)

providesStrongerTrajectoryReadProvides Stronger Trajectory Read(1)

researchPaperIncludesResearch Paper Includes(1)

returnedToReturned to(1)

showedResultsFromShowed Results From(1)

standardEvaluationLengthStandard Evaluation Length(1)

statesAlmostAllDoneStates Almost All Done(1)

warnsUseWithCautionWarns Use With Caution(1)

Other facts (55)

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.

55 facts
PredicateValueRef
IncludeCifar 10[1]
IncludeGem[1]
IncludeFashion Mnist Unpermuted[1]
IncludeFashion Mnist Permuted[1]
IncludeCifar 50[1]
IncludeCifar 5[1]
IncludeCifar 100[1]
IncludeAgem[1]
IncludePacknet[1]
Rdf:typeProject[26]
Rdf:typeProcedure[27]
Rdf:typeModel Evaluation Activity[28]
Ongoingtrue[4]
OngoingLonger Training[13]
Estimated Size100s GB[25]
Estimated SizeTBs[25]
All Have Ci95 Plustrue[1]
Demonstrate SuperiorityCi 95 Plus[1]
Are Parameter Matchedtrue[1]
Include Cnnstrue[1]
Include Ablationstrue[1]
Commit to Improving S2 Beyond ChanceGoal[2]
Requires Largest StorageSnapshots[3]
Presupposes Ongoing ExecutionBackground[5]
Total Estimated Time~25-35 min[5]
Has Iters Per Config20000[5]
Has Config Count4[5]
Running inBackground[5]
Provide Evidence Against Couplingnull[6]
Are Combinations ofModel Variants and Sequence Lengths[7]
Commit to DesignDirection Only[8]
Averaged Over Seedsnull[9]
Short DurationThree K Steps[10]
Exhibit Run to Run VarianceThree K Steps[10]
Nearing CompletionEncoding Comparison[11]
Presupposes Convergence Occurredtrue[12]
Planned Sequentiallynull[14]
Demonstrate CertaintyReplication Success[15]
Commit to Ontological Status ofGelation As Real Phenomenon[15]
Implicates Evolutionary ProcessSymbiogenesis[16]
AfterA Lot of Cleanup[17]
Both Still RunningStatus[18]
Have Normal StatusBpb Best[18]
Illustrate Conceptsnull[19]
On Best Mode PropagatingRootless Suckers Date Stems[19]
Called Fools Experimentstrue[19]
Should ProveGreat Value[20]
Very Successful{}[20]
Gave Satisfactory ResultsThis Country[19]
Should Be Made HereConsidering Climatic Influences[21]
Thorough and Proofefficient[22]
Provide Evidence of FertilityNorthern Plains Victoria[23]
Lead toImprovement[24]
Component TypeSnapshots[25]
Enabled byStep 1 Run Code[28]

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.

55 undated
—
includeblah/papers/part-5
ex:cifar-10
—
allHaveCi95Plusblah/papers/part-5
true
—
demonstrateSuperiorityblah/papers/part-5
ex:ci-95-plus
—
includeblah/papers/part-5
ex:gem
—
includeblah/papers/part-5
ex:fashion-mnist-unpermuted
—
includeblah/papers/part-5
ex:fashion-mnist-permuted
—
includeblah/papers/part-5
ex:cifar-50
—
includeblah/papers/part-5
ex:cifar-5
—
includeblah/papers/part-5
ex:cifar-100
—
includeblah/papers/part-5
ex:agem
—
areParameterMatchedblah/papers/part-5
true
—
includeCnnsblah/papers/part-5
true
—
includeAblationsblah/papers/part-5
true
—
includeblah/papers/part-5
ex:packnet
—
commitToImprovingS2BeyondChanceblah/random/part-38
ex:goal
—
requiresLargestStorageblah/resources/part-50
ex:snapshots
—
ongoingblah/watt-activation/part-1
true
—
presupposesOngoingExecutionblah/watt-activation/part-59
ex:background
—
totalEstimatedTimeblah/watt-activation/part-59
~25-35 min
—
hasItersPerConfigblah/watt-activation/part-59
20000
—
hasConfigCountblah/watt-activation/part-59
4
—
runningInblah/watt-activation/part-59
ex:background
—
provideEvidenceAgainstCouplingblah/watt-activation/part-69
null
—
areCombinationsOfblah/watt-activation/part-72
ex:model-variants-and-sequence-lengths
—
commitToDesignblah/watt-activation/part-281
ex:direction-only
—
averagedOverSeedsblah/watt-activation/part-290
null
—
shortDurationblah/watt-activation/part-289
ex:three-k-steps
—
exhibitRunToRunVarianceblah/watt-activation/part-289
ex:three-k-steps
—
nearingCompletionblah/watt-activation/part-395
ex:encoding-comparison
—
presupposesConvergenceOccurredblah/watt-activation/part-411
true
—
ongoingblah/watt-activation/part-426
ex:longer-training
—
plannedSequentiallyblah/watt-activation/part-451
null
—
demonstrateCertaintyblah/watt-activation/part-468
ex:replication-success
—
commitToOntologicalStatusOfblah/watt-activation/part-468
ex:gelation-as-real-phenomenon
—
implicatesEvolutionaryProcessblah/watt-activation/part-469
ex:symbiogenesis
—
afterblah/watt-activation/part-593
ex:a-lot-of-cleanup
—
bothStillRunningblah/watt-activation/part-690
ex:status
—
haveNormalStatusblah/watt-activation/part-690
ex:bpb-best
—
illustrateConceptstrove-cooktown/mauritius-queensland
null
—
onBestModePropagatingtrove-cooktown/mauritius-queensland
ex:rootless-suckers-date-stems
—
calledFoolsExperimentstrove-cooktown/mauritius-queensland
true
—
shouldProvetrove-cooktown/beche-de-mer
ex:great-value
—
verySuccessfultrove-cooktown/beche-de-mer
{}
—
gaveSatisfactoryResultstrove-cooktown/mauritius-queensland
ex:this-country
—
shouldBeMadeHeretrove-cooktown/john-davis
ex:considering-climatic-influences
—
thoroughAndProoftrove-cooktown/coloured-persons
efficient
—
provideEvidenceOfFertilitybrackenridge-cairns-1880-1900/trove-new/170388541_Wednesday-28-March-1883_No-title
ex:northern-plains-victoria
—
leadToblucher-uhr/wikipedia--killingworth-locomotives
ex:improvement
—
estimatedSizeblah/resources/50
100s GB
—
estimatedSizeblah/resources/50
TBs
—
componentTypeblah/resources/50
ex:snapshots
—
typeblah/watt-activation/64
ex:Project
—
typebeam/11402421-e0dd-4257-81f5-18735667d931
ex:Procedure
—
typebeam/87beddb7-5be9-4b9c-8956-c9ec5a9ce8c0
ex:ModelEvaluationActivity
—
enabledBybeam/87beddb7-5be9-4b9c-8956-c9ec5a9ce8c0
ex:step-1-run-code

References (28)

28 references
  1. [1]Part 514 facts
    ctx:discord/blah/papers/part-5
  2. [2]Part 381 fact
    ctx:discord/blah/random/part-38
  3. [3]Part 501 fact
    ctx:discord/blah/resources/part-50
  4. [4]Part 11 fact
    ctx:discord/blah/watt-activation/part-1
  5. [5]Part 595 facts
    ctx:discord/blah/watt-activation/part-59
  6. [6]Part 691 fact
    ctx:discord/blah/watt-activation/part-69
  7. [7]Part 721 fact
    ctx:discord/blah/watt-activation/part-72
  8. [8]Part 2811 fact
    ctx:discord/blah/watt-activation/part-281
  9. [9]Part 2901 fact
    ctx:discord/blah/watt-activation/part-290
  10. [10]Part 2892 facts
    ctx:discord/blah/watt-activation/part-289
  11. [11]Part 3951 fact
    ctx:discord/blah/watt-activation/part-395
  12. [12]Part 4111 fact
    ctx:discord/blah/watt-activation/part-411
  13. [13]Part 4261 fact
    ctx:discord/blah/watt-activation/part-426
  14. [14]Part 4511 fact
    ctx:discord/blah/watt-activation/part-451
  15. [15]Part 4682 facts
    ctx:discord/blah/watt-activation/part-468
  16. [16]Part 4691 fact
    ctx:discord/blah/watt-activation/part-469
  17. [17]Part 5931 fact
    ctx:discord/blah/watt-activation/part-593
  18. [18]Part 6902 facts
    ctx:discord/blah/watt-activation/part-690
  19. [19]Mauritius Queensland4 facts
    ctx:genes/trove-cooktown/mauritius-queensland
  20. [20]Beche De Mer2 facts
    ctx:genes/trove-cooktown/beche-de-mer
  21. [21]John Davis1 fact
    ctx:genes/trove-cooktown/john-davis
  22. [22]Coloured Persons1 fact
    ctx:genes/trove-cooktown/coloured-persons
  23. [23]170388541 Wednesday 28 March 1883 No Title1 fact
    ctx:genes/brackenridge-cairns-1880-1900/trove-new/170388541_Wednesday-28-March-1883_No-title
  24. [24]Wikipedia Killingworth Locomotives1 fact
    ctx:research/blucher-uhr/wikipedia--killingworth-locomotives
  25. [25]503 facts
    ctx:discord/blah/resources/50
    • full textresources-50
      text/plain3 KBdoc:agent/resources-50/9ddd7f88-cf87-4647-a2bf-984a10ed4b54
      Show excerpt
      [2026-04-07 13:42] xenonfun: yeah getting it more solid. Z.AI was on it all night. Randy also could use and think gonna give it some polish as well. How have some more formal specs on LFS aware Blossom blob storage, storage format, and de…
  26. [26]641 fact
    ctx:discord/blah/watt-activation/64
    • full textwatt-activation-64
      text/plain3 KBdoc:agent/watt-activation-64/620992dd-946e-490a-9a7b-147c1d21d9d4
      Show excerpt
      [2026-03-07 15:38] xenonfun: ``` Excellent results. Comparing against the previous baseline (before gate matmul): ┌───────────────────┬──────────┬────────────────┬─────────────────────────────────┐ │ Config │ Before │ Aft…
  27. [27]11402421 E0dd 4257 81f5 18735667d9311 fact
    ctx:claims/beam/11402421-e0dd-4257-81f5-18735667d931
    • full textbeam-chunk
      text/plain1 KBdoc:beam/11402421-e0dd-4257-81f5-18735667d931
      Show excerpt
      2. **Refine the Search**: If the initial search does not yield significant improvements, consider narrowing down the range or using more sophisticated optimization techniques. 3. **Validate Results**: Validate the results on a separate vali…
  28. [28]87beddb7 5be9 4b9c 8956 C9ec5a9ce8c02 facts
    ctx:claims/beam/87beddb7-5be9-4b9c-8956-c9ec5a9ce8c0
    • full textbeam-chunk
      text/plain1 KBdoc:beam/87beddb7-5be9-4b9c-8956-c9ec5a9ce8c0
      Show excerpt
      ### Next Steps 1. **Run the Provided Code**: Execute the code to experiment with different pre-trained models. 2. **Monitor Results**: Keep an eye on the results and identify the best-performing model. 3. **Report Back**: Once you have the…

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.

↑