INDEX

Explanations

programming terminology and keywords

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

 GenerationType

-0.60

contentLoaded

-0.59

webElementXpaths

-0.54

saraba

-0.53

MessageState

-0.52

extendable

-0.50

Tembelea

-0.49

EDEFAULT

-0.48

onActivityResult

-0.48

DeleteBehavior

-0.47

POSITIVE LOGITS

новништво

0.48

ɵɵelementStart

0.47

ién

0.46

 pinulongan

0.45

tagPool

0.39

likation

0.39

bawa

0.38

étant

0.38

ÁND

0.37

 pungkas

0.37

Activations Density 0.001%

programming terminology and keywords

No Comments

No Known Activations