INDEX

Explanations

references to military service and rank

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

Followed by prepositions, numbers, or titles

np_acts-logits-general · gemini-2.0-flash

military ranks and affiliations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

actionMode

-0.56

 OMITBAD

-0.51

曖昧さ回避

-0.49

krishnan

-0.49

 Supt

-0.48

ี้ย

-0.47

AFTA

-0.47

umbledore

-0.45

cèse

-0.45

 exercice

-0.45

POSITIVE LOGITS

fjspx

0.83

StoryboardSegue

0.65

ellschaft

0.59

ρίες

0.58

 OFDb

0.56

AsUp

0.54

 reserved

0.54

 researched

0.53

Personensuche

0.53

ponses

0.52

Activations Density 0.157%

references to military service and rank

Followed by prepositions, numbers, or titles

military ranks and affiliations

No Comments

No Known Activations