Finetuning
From Dontopedia, the open, paraconsistent wiki. (Last updated 2026-06-06.)
Finetuning has 15 facts recorded in Dontopedia across 6 references.
Mostly:aims at loss reduction(1), is same as pretraining(1), follows pretraining(1)
Maturity scale
raw canonical shape-checked rule-derived certifiedInbound mentions (5)
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.
advocatesGivingUpOnAdvocates Giving Up on(1)
- Lisamegawatts
ex:lisamegawatts
comparesMethodCompares Method(1)
- Message 2
ex:message-2
maySwitchInFinetuneMay Switch in Finetune(1)
- Training Text
ex:training-text
refersToActionRefers to Action(1)
- Log Entry 2026 03 08 04 43
ex:log-entry-2026-03-08-04-43
setsStageForSets Stage for(1)
- Pretraining
ex:pretraining
Other facts (15)
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.
| Predicate | Value | Ref |
|---|---|---|
| Aims at Loss Reduction | null | [1] |
| Is Same As Pretraining | with lower LR | [2] |
| Follows Pretraining | Two Full Passes | [2] |
| Uses Lower Learning Rates | Learning Rates | [2] |
| Yields Better Results Than | Lora | [3] |
| Is Unsuitable Due to | Explosive Synchronization | [4] |
| Rdf:type | Process | [5] |
| Defined by | Log Entry 2026 03 08 04 43 | [5] |
| Has Characteristic | lower learning rates | [5] |
| Adapts Parameter Count | 108000000 | [6] |
| Has Bottleneck | Rank Bottleneck | [6] |
| Allows Specialization | each layer can specialize independently | [6] |
| Specialization Goal | memorize specific facts | [6] |
| Gradient Signal Flow | attention and FFN weights | [6] |
| Has Risk | Catastrophic Forgetting | [6] |
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 (6)
ctx:discord/blah/watt-activation/part-13ctx:discord/blah/watt-activation/part-97ctx:discord/blah/watt-activation/part-173ctx:discord/blah/watt-activation/part-213ctx:discord/blah/watt-activation/97- full textwatt-activation-97text/plain3 KB
doc:agent/watt-activation-97/1b1426ee-68f7-4259-a160-2b0e92e9d3c8Show excerpt
[2026-03-08 04:43] xenonfun: yeah even after 2 full passes I probably have only set the stage for rough understanding of language and then can start finetuning (same thing just lower learning rates and I will possibly switch training text).…
ctx:discord/blah/watt-activation/173- full textwatt-activation-173text/plain3 KB
doc:agent/watt-activation-173/4199481c-aca6-4cd5-a0f3-2b98eb407f8bShow excerpt
[2026-03-09 22:12] xenonfun: ``` === Who is the president of France? === ──────────────────────────────────────────────────────────── K. K. J. / John ──────────────────────────────────────────────────────────── 7 prompt tok (37ms), 8 gene…
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.