INDEX

Explanations

terms related to criminality and suspects

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

criminals, killers, offenders, perpetrators

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

Read

-0.38

chein

-0.38

 Wachs

-0.38

 chees

-0.38

 penny

-0.38

中

-0.38

 guiding

-0.38

Photo

-0.38

Weiter

-0.37

 Penny

-0.36

POSITIVE LOGITS

 offender

0.71

 perpetrator

0.70

 criminals

0.70

 rechange

0.68

 pelaku

0.68

 offenders

0.63

UserScript

0.61

 aggressor

0.61

 feroit

0.60

rungsseite

0.59

Activations Density 0.026%

terms related to criminality and suspects

criminals, killers, offenders, perpetrators

No Comments

No Known Activations