INDEX

Explanations

specific actions or commands related to manipulation or alteration processes

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

 defaultstate

-0.42

 AppModule

-0.33

[]:

-0.31

WithFormat

-0.31

{#

-0.30

ParallelGroup

-0.29

 Polskiej

-0.28

Билгалдахарш

-0.27

曖昧さ回避

-0.27

 peuples

-0.26

POSITIVE LOGITS

webElement

0.67

 indígen

0.66

躇

0.64

fromnode

0.64

protetor

0.61

󠀠

0.61

 erſt

0.61

NameInMap

0.60

 noDo

0.60

 ſich

0.60

Activations Density 0.072%

specific actions or commands related to manipulation or alteration processes

No Comments

No Known Activations