INDEX

Explanations

phrases indicating a refusal to provide comments or information

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

Haunted

-0.72

ãĥİ

-0.65

à¨

-0.64

 Yourself

-0.63

Dro

-0.63

Ranked

-0.62

Â¯Â¯Â¯Â¯

-0.61

nea

-0.60

psc

-0.60

ICAN

-0.59

POSITIVE LOGITS

 comment

1.21

 requests

1.02

 invitations

1.01

 interviews

0.87

 interview

0.85

 repeated

0.83

to

0.81

 inquiries

0.80

 commenting

0.80

 queries

0.80

Activations Density 0.038%

phrases indicating a refusal to provide comments or information

No Comments

No Known Activations