INDEX

Explanations

references to medical information or terms

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

 kasarigan

-0.54

 Tugas

-0.51

 dawned

-0.50

belea

-0.50

SequentialGroup

-0.49

LabelTagHelper

-0.48

jamie

-0.47

JsonPropertyName

-0.47

 gatto

-0.47

 rhestr

-0.47

POSITIVE LOGITS

 poffe

0.41

subsubsection

0.36

liferay

0.34

subsection

0.34

 praw

0.32

 paff

0.31

webElementGuid

0.31

 reaſon

0.30

ümüş

0.29

-/**

0.28

Activations Density 0.007%

references to medical information or terms

No Comments

No Known Activations