INDEX

Explanations

information related to safety and security

oai_token-act-pair · gpt-3.5-turbo

expressions related to safety and secure practices

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Stern

-0.79

imon

-0.62

 Actions

-0.61

 Appearance

-0.60

 response

-0.60

inda

-0.59

ovi

-0.59

 recruitment

-0.58

Ost

-0.57

 Karl

-0.57

POSITIVE LOGITS

 safely

3.95

 securely

1.98

 safe

1.80

 comfortably

1.77

 confidently

1.72

safe

1.67

 efficiently

1.65

 peacefully

1.63

 safest

1.60

 responsibly

1.59

Activations Density 0.012%

information related to safety and security

expressions related to safety and secure practices

No Comments

No Known Activations