INDEX

Explanations

phrases indicating a change in circumstances or outcomes

oai_token-act-pair · gpt-3.5-turbo

phrases indicating change or deterioration in situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

amiya

-0.77

 Reincarn

-0.73

 Preservation

-0.70

ende

-0.67

 Recomm

-0.64

egu

-0.63

esting

-0.63

 Refugees

-0.62

 Maiden

-0.59

Joined

-0.58

POSITIVE LOGITS

 hairy

1.12

 ugly

1.03

 complicated

1.00

 tricky

0.99

 messy

0.98

 worse

0.95

 nasty

0.94

 interesting

0.93

 crazy

0.91

 murky

0.89

Activations Density 0.116%

phrases indicating a change in circumstances or outcomes

phrases indicating change or deterioration in situations

No Comments

No Known Activations