INDEX

Explanations

country names and their associated contextual information.

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

mentions of academic and research institutions in China and medical terms like antibodies and reproductive syndromes.

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

mentions of China or Chinese entities, with particularly strong activation for the word "China" itself.

oai_token-act-pair · claude-3-5-sonnet-20240620 Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.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

}*/

-0.70

🏼

-0.66

 plezier

-0.62

};*/

-0.61

 Winaray

-0.61

 "");

-0.61

}();

-0.60

.)}

-0.60

requisites

-0.60

️

-0.60

POSITIVE LOGITS

parsedMessage

0.82

featureID

0.64

fieldLabel

0.56

पया

0.54

bindParam

0.51

 szóci

0.50

 Rema

0.49

 CHtml

0.49

hideLoading

0.48

protoc

0.48

Activations Density 0.058%

country names and their associated contextual information.

mentions of academic and research institutions in China and medical terms like antibodies and reproductive syndromes.

mentions of China or Chinese entities, with particularly strong activation for the word "China" itself.

No Comments

No Known Activations