INDEX

Explanations

questions and exclamations directed towards someone or a group

oai_token-act-pair · gpt-3.5-turbo

interrogative and exclamatory punctuation marks, particularly question and exclamation marks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

amia

-0.67

Reviewer

-0.66

quartered

-0.65

hran

-0.64

 habit

-0.61

resa

-0.61

cludes

-0.60

 faculties

-0.60

ibles

-0.60

amas

-0.60

POSITIVE LOGITS

/"

1.11

[/

0.94

 Then

0.88

 Later

0.84

etc

0.82

[/

0.80

 referring

0.78

And

0.77

 whereas

0.77

 Well

0.77

Activations Density 0.142%

questions and exclamations directed towards someone or a group

interrogative and exclamatory punctuation marks, particularly question and exclamation marks

No Comments

No Known Activations