INDEX

Explanations

expressions related to questioning and inquiry

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

 <<<<<<<<<<<<<<

-0.66

 saurait

-0.52

 препратки

-0.49

RefNanny

-0.47

 päiv

-0.47

abetes

-0.44

migrationBuilder

-0.43

 MainAxisSize

-0.43

 actualidad

-0.42

genodigd

-0.42

POSITIVE LOGITS

Why

0.68

why

0.67

Why

0.65

 Pourquoi

0.60

WHY

0.59

Varför

0.59

why

0.58

Warum

0.57

 क्यों

0.57

Activations Density 0.064%

expressions related to questioning and inquiry

No Comments

No Known Activations