INDEX

Explanations

instances of security-related actions or measures

oai_token-act-pair · gpt-3.5-turbo

verbs that indicate intention or purpose

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

Li

-0.72

 brim

-0.66

trak

-0.66

Plug

-0.65

 Weld

-0.62

med

-0.61

gins

-0.61

 clipboard

-0.60

 Vest

-0.60

 lineback

-0.59

POSITIVE LOGITS

stakes

0.67

arent

0.65

 nurture

0.64

ANCE

0.64

earances

0.64

iso

0.63

isine

0.62

ASY

0.61

RFC

0.61

 Â©

0.61

Activations Density 0.000%

instances of security-related actions or measures

verbs that indicate intention or purpose

No Comments

No Known Activations