INDEX

Explanations

references to violence and aggression in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 19-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_19/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.19.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

Fprintf

-0.58

ArgsConstructor

-0.57

יצוני

-0.50

strerror

-0.47

clic

-0.44

řád

-0.43

ylus

-0.43

OfDay

-0.43

緒

-0.42

<_>

-0.42

POSITIVE LOGITS

 unsuspecting

1.34

 innocent

1.18

 defen

1.09

innocent

1.00

 targets

0.98

 inocente

0.95

 vulnerable

0.94

 helpless

0.91

 unprotected

0.91

 innoc

0.89

Activations Density 0.593%

references to violence and aggression in various contexts

No Comments

No Known Activations