INDEX

Explanations

names of individuals in various professional fields

oai_token-act-pair · gpt-3.5-turbo

names of individuals and organizations

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

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

 Pigs

-0.78

actionDate

-0.73

Score

-0.70

KO

-0.69

 Saying

-0.66

icol

-0.63

 messing

-0.60

soDeliveryDate

-0.60

 interfering

-0.60

FY

-0.59

POSITIVE LOGITS

 nevertheless

1.06

 boasts

0.97

 opted

0.95

 ventured

0.93

 resorted

0.92

 nonetheless

0.91

 embodies

0.90

 finds

0.90

 drifted

0.90

 struggled

0.89

Activations Density 0.338%

names of individuals in various professional fields

names of individuals and organizations

No Comments

No Known Activations