INDEX

Explanations

content that warns against revealing spoilers in discussions about books

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

Not in Any Lists

No Comments

Negative Logits

ISupport

-0.48

 recensement

-0.41

 Qualification

-0.39

divarius

-0.38

Hart

-0.38

 スキン

-0.37

pipes

-0.37

igneurs

-0.36

Portale

-0.36

Бахар

-0.36

POSITIVE LOGITS

 spoiler

0.65

 reveal

0.63

:✨

0.62

reveal

0.54

 spoilers

0.52

 secret

0.51

 disclose

0.50

 concealing

0.50

 hiding

0.50

 disclosing

0.49

Activations Density 0.281%

content that warns against revealing spoilers in discussions about books

No Comments

No Known Activations