INDEX

Explanations

references to individuals involved in criminal activities or conspiracies

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

persons involved in crimes

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

 المعيارى

-0.53

Super

-0.39

Craig

-0.38

data

-0.36

줌

-0.35

Mega

-0.35

Effective

-0.34

 regioni

-0.34

active

-0.34

 Super

-0.34

POSITIVE LOGITS

 <>",

0.66

IsMutable

0.64

Tembelea

0.57

הערות

0.57

AutoScaleMode

0.56

ImageContext

0.55

tagext

0.53

 ProtoMessage

0.52

 CreateTagHelper

0.51

 iconTwitter

0.51

Activations Density 0.063%

references to individuals involved in criminal activities or conspiracies

persons involved in crimes

No Comments

No Known Activations