INDEX

Explanations

phrases with the word "supposed" followed by a verb

oai_token-act-pair · gpt-3.5-turbo

references to expectations or requirements regarding actions or situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

croft

-0.79

lake

-0.71

isks

-0.71

cloth

-0.69

hold

-0.69

estern

-0.69

bu

-0.68

lust

-0.68

detail

-0.67

Charge

-0.66

POSITIVE LOGITS

 explan

0.90

DonaldTrump

0.88

 mathemat

0.84

 mosqu

0.83

 conflic

0.79

 nodd

0.76

 undermin

0.75

 supposed

0.74

 compe

0.71

 millenn

0.70

Activations Density 0.007%

phrases with the word "supposed" followed by a verb

references to expectations or requirements regarding actions or situations

No Comments

No Known Activations