INDEX

Explanations

elements associated with programming or coding structures

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

EndGlobalSection

-0.47

 purpoſe

-0.46

 ſtate

-0.42

 myſelf

-0.42

 sánchez

-0.41

 ſche

-0.41

ISupport

-0.41

shank

-0.40

 avoient

-0.40

 auffi

-0.40

POSITIVE LOGITS

enumi

0.55

 Савезне

0.52

pushFollow

0.49

expandindo

0.47

 EconPapers

0.47

 MonoBehaviour

0.45

 Autorizaciones

0.44

 ExecuteAsync

0.43

ंदीखरीदारी

0.42

 الرخصة

0.41

Activations Density 1.248%

elements associated with programming or coding structures

No Comments

No Known Activations