INDEX

Explanations

references to outdated or superseded scientific concepts

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

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_1m/average_l0_107

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

1,048,576

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

𝘂

-0.78

𝗿

-0.76

𝘃

-0.75

𝘅

-0.72

𝗯

-0.69

𝘇

-0.69

𝗽

-0.67

𝗻

-0.67

#![

-0.65

lawayo

-0.65

POSITIVE LOGITS

 Astrology

1.49

 astrology

1.45

 pinulongan

0.94

 Madura

0.84

 iſt

0.82

Efq

0.75

 itſelf

0.74

 dieß

0.73

).)

0.73

 ―――――

0.73

Activations Density 0.000%

references to outdated or superseded scientific concepts

No Comments

No Known Activations