INDEX

Explanations

phrases related to personal identity and social interactions

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

BeginInit

-0.50

ymce

-0.47

 ModelRenderer

-0.46

PerformLayout

-0.46

 useRouter

-0.45

AnchorStyles

-0.45

consulté

-0.45

astify

-0.45

ConstraintMaker

-0.44

 meille

-0.43

POSITIVE LOGITS

 instead

0.45

instead

0.40

ArgsConstructor

0.40

uttosto

0.37

反而

0.36

 Instead

0.36

 merely

0.36

 rather

0.35

 Rather

0.35

 misinterpreted

0.35

Activations Density 0.066%

phrases related to personal identity and social interactions

No Comments

No Known Activations