INDEX

Explanations

Twitter handles and mentions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_11/width_16k/average_l0_80

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.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

abestanden

-0.99

 ―――――

-0.93

DeleteBehavior

-0.91

HtmlAttribute

-0.87

 Shaksp

-0.85

WindowConstants

-0.83

 препратки

-0.81

amaño

-0.79

rawDesc

-0.78

XMLSchema

-0.78

POSITIVE LOGITS

0.86

0.69

'@

0.55

(@

0.54

'@

0.52

"@

0.51

في

0.47

,@

0.47

"@

0.47

 desped

0.47

Activations Density 0.152%

Twitter handles and mentions

No Comments

No Known Activations