INDEX

Explanations

references to clocks and time-related phrases

oai_token-act-pair · gpt-3.5-turbo

references to clocks and time-related concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Flavoring

-0.99

Reviewer

-0.81

Urban

-0.79

BuyableInstoreAndOnline

-0.79

Priv

-0.77

rito

-0.76

Ô

-0.76

qua

-0.75

 Nicarag

-0.73

ortium

-0.72

POSITIVE LOGITS

 clock

1.10

clock

1.01

 Clock

1.00

 clocks

0.97

work

0.95

 ticking

0.95

 ticks

0.88

tower

0.88

wise

0.86

 tower

0.80

Activations Density 0.014%

references to clocks and time-related phrases

references to clocks and time-related concepts

No Comments

No Known Activations