Post Training Ternary Quantization

as donto believed on

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

Post Training Ternary Quantization has 5 facts recorded in Dontopedia across 2 references, with 1 live disagreement.

5 facts·4 predicates·2 sources·1 in dispute

Mostly:would destroy(2), is destructive(1), rdf:type(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.

5 facts

Predicate	Value	Ref
Would Destroy	This Model	[1]
Would Destroy	Ternary Quantization Model	[2]
Is Destructive	Lohe Model	[1]
Rdf:type	Quantization Strategy	[2]
Is Wrong Approach	true	[2]

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.

5 undated

—

isDestructiveblah/watt-activation/part-277

ex:lohe-model

—

wouldDestroyblah/watt-activation/part-277

ex:this-model

—

typeblah/watt-activation/275

ex:QuantizationStrategy

—

wouldDestroyblah/watt-activation/275

ex:ternary-quantization-model

—

isWrongApproachblah/watt-activation/275

true

References (2)

2 references

[1]Part 2772 facts
ctx:discord/blah/watt-activation/part-277
[2]2753 facts
ctx:discord/blah/watt-activation/275
- full textwatt-activation-275
  text/plain3 KBdoc:agent/watt-activation-275/149bf24b-54fb-4412-b289-e2e03bccffe2
  Show excerpt
  [2026-03-13 20:46] xenonfun: ⏺ This tells a clear story: Current trained weights are NOT naturally ternary. Only ~46% of values are near {-1, 0, 1} after scaling (random Gaussian would give ~47%, so no better than chance). Relative err…

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.

Post Training Ternary Quantization

Other facts (5)

Timeline

References (2)

See also

Keep researching