INDEX

Explanations

quotations with emotionally charged phrases

oai_token-act-pair · gpt-3.5-turbo

phrases that end with a comma, indicating lists or ongoing dialogue

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

etheless

-0.95

»Ĵ

-0.80

erity

-0.74

İĭ

-0.72

¬¼

-0.70

anuts

-0.69

ĵĺ

-0.67

thur

-0.66

£ı

-0.66

ãĥĳ

-0.65

POSITIVE LOGITS

 says

1.12

 said

1.08

 reads

1.04

said

1.03

 recalls

0.97

 commented

0.90

 wrote

0.88

 joked

0.87

 explains

0.82

 remarked

0.82

Activations Density 0.087%

quotations with emotionally charged phrases

phrases that end with a comma, indicating lists or ongoing dialogue

No Comments

No Known Activations