INDEX

Explanations

actions and concepts related to separation or division

oai_token-act-pair · gpt-3.5-turbo

instances of the word "spar."

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

ĪĴ

-0.70

ħĭ

-0.69

bern

-0.67

BOOK

-0.63

hirt

-0.63

sburg

-0.62

 Brill

-0.62

ĨĴ

-0.61

ILE

-0.61

ĸļ

-0.60

POSITIVE LOGITS

allel

1.58

liament

1.33

anoia

1.23

alysis

1.02

ities

0.97

ity

0.93

imony

0.93

vati

0.92

amount

0.90

agraph

0.90

Activations Density 0.031%

actions and concepts related to separation or division

instances of the word "spar."

No Comments

No Known Activations