INDEX

Explanations

words indicating the suitability or acceptability of concepts, plans, or situations

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

ViewFeatures

-0.78

ScopeManager

-0.63

 wireType

-0.62

TagHelpers

-0.62

ьаж

-0.61

RTLR

-0.60

RTEE

-0.58

 ***!

-0.57

WidgetItem

-0.57

ConstraintMaker

-0.56

POSITIVE LOGITS

or

0.62

或

0.49

หรือ

0.48

 Expédié

0.47

 или

0.46

 maupun

0.44

and

0.42

 Gegenwart

0.41

 Griechen

0.40

Activations Density 0.218%

words indicating the suitability or acceptability of concepts, plans, or situations

No Comments

No Known Activations