INDEX

Explanations

mentions of the term "Hopkins."

oai_token-act-pair · gpt-3.5-turbo

mentions of "Hopkins" and "Johns Hopkins University."

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

gered

-0.84

cing

-0.78

regate

-0.70

gers

-0.68

ffen

-0.66

pt

-0.66

$$$$

-0.66

razil

-0.66

ãĥł

-0.65

opsis

-0.62

POSITIVE LOGITS

 Hopkins

1.18

hip

1.01

kins

0.99

kinson

0.99

bay

0.83

ley

0.81

erenn

0.80

stone

0.80

endish

0.79

 Hutch

0.76

Activations Density 0.024%

mentions of the term "Hopkins."

mentions of "Hopkins" and "Johns Hopkins University."

No Comments

No Known Activations