INDEX

Explanations

conversational expressions and inquiries into personal background and identity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 39-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_39/width_16k/average_l0_131

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.39.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

 Italij

-0.79

 Grecs

-0.77

 Einfluß

-0.71

 Meksika

-0.70

 Inscrivez

-0.65

 plufieurs

-0.64

 Sünde

-0.62

 Füßen

-0.62

 myſelf

-0.61

 Bewußt

-0.61

POSITIVE LOGITS

0.70

0.56

col

0.56

phi

0.56

 são

0.55

 hanno

0.55

0.54

 hemel

0.53

 aber

0.53

 miss

0.53

Activations Density 1.018%

conversational expressions and inquiries into personal background and identity

No Comments

No Known Activations