INDEX

Explanations

phrases that indicate frequency or probability

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate frequency or typicality in actions or statements

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

andering

-0.75

imm

-0.71

enting

-0.67

driving

-0.64

çĶ

-0.62

 rising

-0.61

 gold

-0.61

ä½ľ

-0.60

ander

-0.59

å®

-0.59

POSITIVE LOGITS

entimes

1.22

terness

1.02

etheless

0.97

yip

0.95

veyard

0.91

eus

0.85

 Called

0.85

ccording

0.84

asionally

0.83

 nomine

0.81

Activations Density 0.012%

phrases that indicate frequency or probability

phrases that indicate frequency or typicality in actions or statements

No Comments

No Known Activations