INDEX

Explanations

proper nouns related to titles, organizations, and geographical features

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 31-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_31/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.31.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

axel

-0.49

 ویکی‌پدی

-0.47

Tikang

-0.44

 незавершена

-0.44

InteropServices

-0.43

 հղումներ

-0.42

setCellStyle

-0.42

tomans

-0.41

%</

-0.41

GOTREF

-0.41

POSITIVE LOGITS

awtextra

0.48

 '\\;'

0.40

bcryptjs

0.39

 Putih

0.39

 betweenstory

0.38

EndProject

0.36

évaluateur

0.36

 Vikipedi

0.35

rungsseite

0.34

новништво

0.34

Activations Density 0.398%

proper nouns related to titles, organizations, and geographical features

No Comments

No Known Activations