INDEX

Explanations

words related to authorship and convenience

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_0/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

parsedMessage

-0.98

 pleaſure

-0.81

 Signalez

-0.69

TintMode

-0.66

 CreateTagHelper

-0.65

majánló

-0.63

 Anſ

-0.62

 Houſe

-0.59

 ſind

-0.59

Gaia

-0.59

POSITIVE LOGITS

 adequate

0.58

 AUTHOR

0.57

 authorized

0.57

author

0.57

 authorizing

0.56

 authorize

0.54

 conveniently

0.54

 convenient

0.50

 South

0.48

 author

0.48

Activations Density 0.314%

words related to authorship and convenience

No Comments

No Known Activations