INDEX

Explanations

the word "watch" and its variations in relation to viewing content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

Abp

-0.81

/>";

-0.78

cycline

-0.71

 للمعارف

-0.71

}}],

-0.71

(;;)

-0.70

])),

-0.70

']),

-0.68

complexContent

-0.67

gdala

-0.67

POSITIVE LOGITS

 watch

2.38

 Watch

2.34

 WATCH

2.26

 watches

2.17

watch

2.15

Watch

2.15

WATCH

2.04

 Watches

1.96

watches

1.91

 watched

1.89

Activations Density 0.107%

the word "watch" and its variations in relation to viewing content

No Comments

No Known Activations