INDEX

Explanations

references to money laundering and political corruption

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

political figures and names

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

jeka

-0.31

DataBinding

-0.30

 cober

-0.30

 fully

-0.29

 acol

-0.29

≔

-0.29

 suited

-0.28

 форма

-0.28

nasium

-0.28

⎠

-0.27

POSITIVE LOGITS

BASEPATH

0.66

ckså

0.58

 &___

0.57

msgTypes

0.56

ViewFeatures

0.54

 ProtoMessage

0.53

 disambiguazione

0.52

 interlocutor

0.52

 שוליים

0.52

tempts

0.51

Activations Density 0.065%

references to money laundering and political corruption

political figures and names

No Comments

No Known Activations