INDEX

Explanations

names or entities with 'Gr' followed by a single digit, specifically 'Gr9'

oai_token-act-pair · gpt-3.5-turbo

references to the word "Grady"

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

eers

-0.73

manship

-0.67

 shortened

-0.67

doors

-0.66

quo

-0.65

 pred

-0.65

 messenger

-0.63

 focal

-0.62

peak

-0.62

)=(

-0.62

POSITIVE LOGITS

acies

0.97

asp

0.95

izont

0.93

gr

0.92

ands

0.91

ains

0.91

igs

0.91

itt

0.91

iffin

0.90

umpy

0.89

Activations Density 0.005%

names or entities with 'Gr' followed by a single digit, specifically 'Gr9'

references to the word "Grady"

No Comments

No Known Activations