INDEX

Explanations

warnings about potential dangers and accidents

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 fufficient

-0.42

 mont

-0.41

 partic

-0.41

stalt

-0.40

+:+

-0.38

pushFollow

-0.38

addGap

-0.38

 cooper

-0.38

天下

-0.38

 autorytatywna

-0.38

POSITIVE LOGITS

 recurrir

0.54

 resorted

0.51

 temptation

0.50

 resorting

0.46

číta

0.44

NameInMap

0.43

 tempted

0.42

 cenderung

0.42

ActionCreators

0.41

 Préférences

0.41

Activations Density 0.427%

warnings about potential dangers and accidents

No Known Activations