INDEX

Explanations

references to specific media outlets and publications

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 nakalista

-0.58

intios

-0.54

exitRule

-0.53

حياتها

-0.46

 AssemblyCulture

-0.46

posedge

-0.45

RetentionPolicy

-0.45

WithIOException

-0.45

Diweddarwch

-0.44

巯

-0.43

POSITIVE LOGITS

 Huffington

0.66

存于互联网档案馆

0.60

UserScript

0.52

huff

0.52

 bloggers

0.50

astify

0.49

 Newsweek

0.48

 blogs

0.48

 חיצוניים

0.47

ypress

0.45

Activations Density 0.017%

references to specific media outlets and publications

No Comments

No Known Activations