Dontopedia

G-Alpha

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

G-Alpha has 24 facts recorded in Dontopedia across 4 references, with 3 live disagreements.

24 facts·17 predicates·4 sources·3 in dispute

Mostly:uses activation(4), has loss(2), loss recorded at step(2)

Maturity scale raw canonical shape-checked rule-derived certified

Inbound mentions (3)

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.

firstCandidateFirst Candidate(1)

hasParentHas Parent(1)

isMutationOfIs Mutation of(1)

Other facts (22)

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.

22 facts
PredicateValueRef
Uses ActivationGelu[1]
Uses ActivationGelu[2]
Uses ActivationGelu Activation[3]
Uses ActivationGelu[4]
Has Loss8.37[4]
Has Loss7.92[4]
Loss Recorded at Step0[4]
Loss Recorded at Step50[4]
Has Loss Drop From8.37[1]
Has Loss to7.92[1]
Is First Candidatenull[1]
Demonstrates Quick Progressnull[1]
Achieves Token Speed7400[1]
Swaps to Next Candidate at Step3000[2]
Has Completed Steps30[2]
Swap Time Approx41 min[2]
Leads Current Progresstrue[2]
References Prior Contexttrue[2]
Rdf:typeAgent[3]
Finished Steps30[3]
Is Instance ofGeneration 0[4]
Loss Decreasingtrue[4]

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.

hasLossDropFromblah/training-and-evals/part-22
8.37
hasLossToblah/training-and-evals/part-22
7.92
isFirstCandidateblah/training-and-evals/part-22
null
demonstratesQuickProgressblah/training-and-evals/part-22
null
achievesTokenSpeedblah/training-and-evals/part-22
7400
usesActivationblah/training-and-evals/part-22
ex:gelu
swapsToNextCandidateAtStepblah/training-and-evals/part-23
3000
hasCompletedStepsblah/training-and-evals/part-23
30
usesActivationblah/training-and-evals/part-23
ex:gelu
swapTimeApproxblah/training-and-evals/part-23
41 min
leadsCurrentProgressblah/training-and-evals/part-23
true
referencesPriorContextblah/training-and-evals/part-23
true
typeblah/training-and-evals/23
ex:Agent
labelblah/training-and-evals/23
G-Alpha
usesActivationblah/training-and-evals/23
ex:gelu-activation
finishedStepsblah/training-and-evals/23
30
labelblah/training-and-evals/22
G-Alpha
usesActivationblah/training-and-evals/22
ex:gelu
hasLossblah/training-and-evals/22
8.37
hasLossblah/training-and-evals/22
7.92
lossRecordedAtStepblah/training-and-evals/22
0
lossRecordedAtStepblah/training-and-evals/22
50
isInstanceOfblah/training-and-evals/22
ex:generation-0
lossDecreasingblah/training-and-evals/22
true

References (4)

4 references
  1. [1]Part 226 facts
    ctx:discord/blah/training-and-evals/part-22
  2. [2]Part 236 facts
    ctx:discord/blah/training-and-evals/part-23
  3. [3]234 facts
    ctx:discord/blah/training-and-evals/23
    • full texttraining-and-evals-23
      text/plain3 KBdoc:agent/training-and-evals-23/70bec41a-b6db-4999-9455-15d7176b4205
      Show excerpt
      [2026-02-25 14:27] omega [bot]: **Symbio Search Math Problem** Current run `super_chat_20260225140824_ee4k` has a config mismatch: ``` 50000 candidates × 3000 steps × 20 gens = 3,000,000,000 total search steps But only 50,000 total trainin
  4. [4]228 facts
    ctx:discord/blah/training-and-evals/22
    • full texttraining-and-evals-22
      text/plain2 KBdoc:agent/training-and-evals-22/c7c2c5ec-1562-4f65-85b1-b3e9028e864e
      Show excerpt
      [2026-02-25 14:22] omega [bot]: **Evolutionary Search Improvements** - **Lineage tracking**: Every candidate now has a unique ID, name, parent ID, and full lineage array. Names use Greek letter system — Gen 0: `G-Alpha`, `S-Beta`; mutations

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.