INDEX

Explanations

comparative adjectives indicating size and quality

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.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

mit

-0.70

ymce

-0.67

PerformLayout

-0.64

Bad

-0.64

Ben

-0.63

God

-0.63

 Turnbull

-0.61

Mat

-0.61

URIComponent

-0.61

Get

-0.60

POSITIVE LOGITS

 leſs

1.30

 than

1.23

 myſelf

1.18

 clearer

1.15

theless

1.13

 wiser

1.13

 himſelf

1.10

 greener

1.09

 Anſ

1.09

ſelf

1.09

Activations Density 0.076%

comparative adjectives indicating size and quality

No Comments

No Known Activations