INDEX

Explanations

dates represented in the year 2000

oai_token-act-pair · gpt-3.5-turbo

occurrences of the year 2000 and references associated with that time period

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

warts

-0.87

atories

-0.76

Else

-0.74

unh

-0.70

attled

-0.67

unal

-0.66

alth

-0.65

iqu

-0.65

artifacts

-0.65

atory

-0.64

POSITIVE LOGITS

mAh

0.86

 Leban

0.79

istar

0.75

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

0.71

 onwards

0.69

 Reincarn

0.69

Hz

0.69

enaries

0.69

rpm

0.69

visors

0.68

Activations Density 0.013%

dates represented in the year 2000

occurrences of the year 2000 and references associated with that time period

No Comments

No Known Activations