INDEX

Explanations

numerical values or references

oai_token-act-pair · gpt-4o-mini Triggered by @bot

evidentiary assertions within legal contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

legal filings and arguments

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 AspNetCore

-0.55

 Biôgrafia

-0.52

RunWith

-0.52

 FALL

-0.51

îtra

-0.51

erek

-0.50

фаль

-0.50

Gar

-0.49

 ModelExpression

-0.49

 nawr

-0.48

POSITIVE LOGITS

存于互联网档案馆

0.63

 mute

0.58

 mergeFrom

0.57

Parcelize

0.55

 виправивши

0.55

 Houſe

0.55

 Puedes

0.53

 ſeveral

0.53

Suf

0.53

Javascript

0.52

Activations Density 0.020%

numerical values or references

evidentiary assertions within legal contexts

legal filings and arguments

No Comments

No Known Activations