INDEX

Explanations

references to characters or individuals, particularly in storytelling or narrative contexts

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

UrlResolution

-0.47

RectangleBorder

-0.44

 AssemblyCompany

-0.43

StoryboardSegue

-0.41

 mijne

-0.40

̈́

-0.39

 GenerationType

-0.39

DECREF

-0.39

 ویکی‌پدی

-0.38

insegna

-0.37

POSITIVE LOGITS



0.54

:✨

0.50

 betweenstory

0.47

 Shukla

0.47

Brazil

0.46

 Yadav

0.45

 Rast

0.43

Gai

0.43

Postal

0.43

madu

0.43

Activations Density 0.287%

references to characters or individuals, particularly in storytelling or narrative contexts

No Comments

No Known Activations