INDEX

Explanations

references to readers or audience engagement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_3/width_16k/average_l0_90

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.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

########.

-0.50

Autowired

-0.40

 protoimpl

-0.35

NewUrlParser

-0.34

 kaarangay

-0.34

 quotes

-0.34

Bakgrunnsstoff

-0.33

PROMOTED

-0.32

 bills

-0.32

 Bill

-0.31

POSITIVE LOGITS

 readers

2.61

 viewers

2.53

 listeners

2.16

 audiences

2.13

 Readers

2.08

Readers

1.95

 viewer

1.94

 audience

1.87

 reader

1.85

readers

1.79

Activations Density 0.180%

references to readers or audience engagement

No Comments

No Known Activations