INDEX

Explanations

numerical figures or statistics

oai_token-act-pair · gpt-3.5-turbo

numerical values or counts related to various topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

swe

-0.69

 creature

-0.66

 bear

-0.64

 cens

-0.63

 lifes

-0.62

 tongues

-0.62

axe

-0.61

 trump

-0.61

 tram

-0.60

 unconditional

-0.60

POSITIVE LOGITS

Advertisement

1.23

Meanwhile

1.22

However

1.20

Besides

1.18

Also

1.16

Similarly

1.16

Likewise

1.15

Some

1.14

Related

1.14

Both

1.14

Activations Density 0.673%

numerical figures or statistics

numerical values or counts related to various topics

No Comments

No Known Activations