INDEX

Explanations

phrases related to specific time periods or events

oai_token-act-pair · gpt-3.5-turbo

references to time, specifically hours

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

umo

-0.71

PsyNetMessage

-0.70

ictions

-0.70

usal

-0.68

ifiable

-0.67

 discharge

-0.67

ulatory

-0.66

ulated

-0.66

iction

-0.66

uton

-0.65

POSITIVE LOGITS

Ago

1.15

terness

1.02

 Hour

1.01

 Lauder

0.94

chen

0.81

DAQ

0.77

 ILCS

0.75

 Hours

0.75

 Wife

0.73

cade

0.73

Activations Density 0.013%

phrases related to specific time periods or events

references to time, specifically hours

No Comments

No Known Activations