INDEX

Explanations

conditional phrases and conjunctions indicating possibilities or choices

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 23-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_23/width_16k/average_l0_75

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.23.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Diwedd

-0.63

InstrumentedTest

-0.61

ướng

-0.56

 henvisninger

-0.54

bardier

-0.51

UnusedPrivate

-0.51

})));

-0.50

мента

-0.49

 Préférences

-0.49

Vidite

-0.48

POSITIVE LOGITS

addCriterion

0.56

 CreateTagHelper

0.55

horabuena

0.55

 leçon

0.52

 sostanza

0.51

 avven

0.50

 mergeFrom

0.50

 importantly

0.50

 förb

0.50

 useStyles

0.50

Activations Density 0.227%

conditional phrases and conjunctions indicating possibilities or choices

No Comments

No Known Activations