INDEX

Explanations

mentions and descriptions related to workplace activities

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "work."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Ukrain

-0.92

olic

-0.72

 Russo

-0.69

Sov

-0.69

 wholesale

-0.65

ACP

-0.63

emonic

-0.62

Bol

-0.61

laus

-0.61

pha

-0.60

POSITIVE LOGITS

aday

1.13

station

1.08

bench

1.00

 ethic

0.97

flows

0.97

manship

0.93

night

0.92

days

0.89

day

0.89

lifting

0.87

Activations Density 0.057%

mentions and descriptions related to workplace activities

references to the concept of "work."

No Comments

No Known Activations