INDEX

Explanations

instances of the word "refuse" and its variations, indicating rejection or denial

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

InputBorder

-0.68

contentPadding

-0.54

bootstrapcdn

-0.53

tagHelper

-0.51

わけで

-0.50

RectangleBorder

-0.49

Johann

-0.49

YG

-0.47

 isle

-0.45

TT

-0.44

POSITIVE LOGITS

 refuse

0.75

 Refuse

0.70

 refused

0.68

 refuses

0.67

 refusal

0.67

 refus

0.62

 refusing

0.58

 refuser

0.57

 rifi

0.57

 rechazo

0.56

Activations Density 0.012%

instances of the word "refuse" and its variations, indicating rejection or denial

No Comments

No Known Activations