INDEX

Explanations

mentions or references to social media metrics like followers, views, likes, and comments

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

fug

-0.74

ropri

-0.71

Alz

-0.70

 insol

-0.68

 manslaughter

-0.68

EH

-0.68

 realism

-0.66

 Quart

-0.65

 Shell

-0.65

 Goddard

-0.64

POSITIVE LOGITS

hip

1.14

hips

0.95

hare

0.91

Tweet

0.87

ularity

0.84

Reviewer

0.84

cats

0.83

 Likes

0.81

iences

0.81

pages

0.81

Activations Density 0.185%

mentions or references to social media metrics like followers, views, likes, and comments

No Comments

No Known Activations