10K checkpoint
From Dontopedia, the open, paraconsistent wiki. (Last updated 2026-06-06.)
10K checkpoint has 6 facts recorded in Dontopedia across 3 references.
Mostly:features(1), has qualitative marker(1), precedes(1)
Maturity scale
raw canonical shape-checked rule-derived certifiedInbound 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.
comparesCheckpointCompares Checkpoint(1)
- Inference Comparison 30k
ex:inference-comparison-30k
locatedInLocated in(1)
- Restored Optimizer State
ex:restored-optimizer-state
usedCheckpointUsed Checkpoint(1)
- Forward Pass 1
ex:forward-pass-1
Other facts (5)
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 |
|---|---|---|
| Features | Word Shapes Only | [1] |
| Has Qualitative Marker | word-shapes only, fragment tokens ("spronentive", "mitalogical") | [1] |
| Precedes | Checkpoint 20k | [1] |
| Produces | Fragment Tokens | [1] |
| Has Iteration Count | 10000 | [2] |
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 (3)
ctx:discord/blah/watt-activation/part-663ctx:discord/blah/watt-activation/91- full textwatt-activation-91text/plain3 KB
doc:agent/watt-activation-91/61aeb5f1-b837-4123-a6c4-b76d719d0b25Show excerpt
[2026-03-08 00:46] xenonfun: ``` Looks good. Now let me calculate the exact iterations and launch: - 133M tokens / 4096 = 32,471 iters per epoch - 2 epochs = 64,942 total iters - Already done: 10,000 - Remaining: 54,942 iters (rou…
ctx:discord/blah/watt-activation/138- full textwatt-activation-138text/plain3 KB
doc:agent/watt-activation-138/9bd08756-4bea-4055-bcfd-5d3ed62434a1Show excerpt
[2026-03-09 06:23] xenonfun: ``` ⏺ No — softmax is on self.log_adjacency which is a static (G, G) parameter tensor, completely independent of T. It runs once per forward pass in O(G²) = O(64). The sequence-length work is entirely in _gate…
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.