INDEX

Explanations

references to nondescript or unspecified entities

oai_token-act-pair · gpt-3.5-turbo

terms related to non-discrimination and non-toxic classifications

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

Package

-0.92

ppa

-0.77

 Syndicate

-0.69

================

-0.67

 Majesty

-0.66

 Divide

-0.64

gs

-0.64

 Royale

-0.63

 Slug

-0.63

 Clip

-0.63

POSITIVE LOGITS

etheless

1.07

oubted

1.00

 nond

0.97

oub

0.92

minist

0.91

ouble

0.91

airy

0.89

roit

0.87

uce

0.86

aily

0.86

Activations Density 0.012%

references to nondescript or unspecified entities

terms related to non-discrimination and non-toxic classifications

No Comments

No Known Activations