INDEX

Explanations

dates formatted in the day-month-year structure

oai_token-act-pair · gpt-3.5-turbo

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

 Genie

-0.59

 hobbies

-0.55

 explan

-0.54

 expectations

-0.53

 streng

-0.53

DN

-0.51

 habitual

-0.51

ilater

-0.50

itton

-0.50

 Alban

-0.50

POSITIVE LOGITS

th

1.87

rd

1.33

ths

1.11

teenth

1.11

nd

1.10

TH

1.06

tha

1.01

st

0.95

thus

0.93

thal

0.88

Activations Density 0.084%

dates formatted in the day-month-year structure

No Comments

No Known Activations