INDEX

Explanations

negative statements or denials

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

ValueStyle

-0.50

ViewImports

-0.47

PerformLayout

-0.42

webElementGuid

-0.41

.*")]

-0.41

 Pears

-0.41

WithFormat

-0.38

ulipas

-0.38

gende

-0.37

useEffect

-0.35

POSITIVE LOGITS

 necessarily

0.59

 tampoco

0.55

 nevy

0.50

Reprodução

0.49

necessarily

0.47

 Walkover

0.47

 keineswegs

0.46

 niestety

0.46

決して

0.46

 אינו

0.45

Activations Density 0.052%

negative statements or denials

No Comments

No Known Activations