INDEX

Explanations

phrases or sentences where the speaker expresses a strong emotional reaction or opinion

oai_token-act-pair · gpt-3.5-turbo

the word "like" used excessively to imply comparison or emphasis

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

verning

-0.83

hiba

-0.74

Äĩ

-0.74

ouri

-0.73

anthrop

-0.71

alogue

-0.71

iche

-0.70

etts

-0.69

doi

-0.67

Ã¨re

-0.66

POSITIVE LOGITS

lihood

1.07

lier

0.81

 minded

0.76

 parity

0.74

liest

0.70

 crazy

0.70

 drafts

0.68

terday

0.67

liness

0.66

ably

0.65

Activations Density 0.083%

phrases or sentences where the speaker expresses a strong emotional reaction or opinion

the word "like" used excessively to imply comparison or emphasis

No Comments

No Known Activations