INDEX

Explanations

references to communication and non-verbal expression

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

SequentialGroup

-0.59

 المعيارى

-0.52

bcryptjs

-0.49

enterOuterAlt

-0.49

 newOwner

-0.46

 الرياضيه

-0.46

 houſe

-0.44

addComponent

-0.44

__(/*!

-0.44

ſelf

-0.42

POSITIVE LOGITS

 nonverbal

0.52

 silent

0.51

 communicate

0.48

تقاوى

0.47

 silently

0.46

silent

0.45

RetentionPolicy

0.43

 communicates

0.42

无声

0.41

 silen

0.40

Activations Density 0.354%

references to communication and non-verbal expression

No Comments

No Known Activations