INDEX

Explanations

statements involving official announcements or reports from various sources

oai_token-act-pair · gpt-3.5-turbo

references to legal and regulatory matters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 misconception

-0.57

 misunderstanding

-0.53

ogg

-0.53

 Magical

-0.53

iatus

-0.52

 QUEST

-0.52

 misunderstand

-0.51

ipeg

-0.50

ctive

-0.50

ukong

-0.49

POSITIVE LOGITS

 respectively

0.85

enegger

0.67

anwhile

0.63

éĥ

0.63

actionDate

0.59

 thereby

0.59

bourg

0.57

 jointly

0.56

 reportedly

0.55

]."

0.54

Activations Density 3.286%

statements involving official announcements or reports from various sources

references to legal and regulatory matters

No Comments

No Known Activations