INDEX

Explanations

the word "subsequent" and similar variations related to subsequent actions or events

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

 броса

-0.43

참고

-0.40

Directive

-0.40

 Chau

-0.39

 Newspapers

-0.38

 Faul

-0.38

öszi

-0.38

sendStatus

-0.38

 repeated

-0.38

 watches

-0.37

POSITIVE LOGITS

 subsequent

0.74

principalColumn

0.73

 iron

0.71

Cu

0.70

Cu

0.69

 Iron

0.68

 copper

0.66

Iron

0.66

 IRON

0.66

 EconPapers

0.66

Activations Density 0.167%

the word "subsequent" and similar variations related to subsequent actions or events

No Comments

No Known Activations