INDEX

Explanations

mentions of specific dates and events

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "the"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 teasp

-0.80

 besides

-0.76

pots

-0.75

zac

-0.72

olt

-0.70

arettes

-0.70

atown

-0.70

thood

-0.68

verage

-0.67

emption

-0.66

POSITIVE LOGITS

 Department

1.00

 Associated

0.99

 Supreme

0.95

 United

0.95

 venerable

0.95

 Ministry

0.90

 Institute

0.89

 Securities

0.89

 Bureau

0.89

 Centers

0.88

Activations Density 0.270%

mentions of specific dates and events

occurrences of the word "the"

No Comments

No Known Activations