Gelu Mlp Baseline
From Dontopedia, the open, paraconsistent wiki. (Last updated 2026-06-06.)
Gelu Mlp Baseline has 13 facts recorded in Dontopedia across 2 references, with 1 live disagreement.
Mostly:wins on quality(2), has bpb at500(1), has most params in ffn(1)
Maturity scale
raw canonical shape-checked rule-derived certifiedWins on Qualityin disputewinsOnQuality
- Ffn Comparison Table 1[2]all time · Part 386
- true[1]sourceall time · 384
Has Bpb At500hasBpbAt500
- 2.334[2]all time · Part 386
Has Most Params in FfnhasMostParamsInFfn
- Ffn Comparison Table 1[2]all time · Part 386
Has Tokens Per SecondhasTokensPerSecond
- 31000[2]all time · Part 386
Has Total Params ApproxhasTotalParamsApprox
- 12M MLP heavy[2]all time · Part 386
Quality ReasonqualityReason
- has the most parameters in the FFN[1]sourceall time · 384
Model DescriptionmodelDescription
- MLP heavy[1]sourceall time · 384
Params Total ModelparamsTotalModel
- ~12M[1]sourceall time · 384
Tokens Per SecondtokensPerSecond
- 31K[1]sourceall time · 384
Bpb At500bpbAt500
- 2.334[1]sourceall time · 384
Rdfs:labelrdfs:label
- gelu (MLP baseline)[1]sourceall time · 384
Rdf:typerdf:type
Inbound mentions (4)
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.
areFasterThanAre Faster Than(1)
- Manifold Methods
ex:manifold-methods
comparesFFNTypeCompares Ffn Type(1)
- Ffn Comparison Table 1
ex:ffn-comparison-table-1
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.
References (2)
- custom
ctx:discord/blah/watt-activation/384- full textwatt-activation-384text/plain3 KB
doc:agent/watt-activation-384/9e57a6f6-7f88-4fd5-81f0-454435de5e3eShow excerpt
[2026-03-19 01:53] xenonfun: ⏺ Pure FFN comparison — clean, no contention, sequential: ``` ┌─────────────────────┬──────────┬───────┬──────────────────────┐ │ FFN Type │ BPB @500 │ tok/s │ Params (total model) │ ├──────────…
- custom
ctx:discord/blah/watt-activation/part-386
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.