INDEX

Explanations

and highlight corrections or amendments in a text

oai_token-act-pair · gpt-3.5-turbo

instances of the word "correction" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

atos

-0.82

NetMessage

-0.72

ramid

-0.69

otor

-0.69

atom

-0.69

lust

-0.64

aden

-0.63

soDeliveryDate

-0.62

axy

-0.62

 Valhalla

-0.61

POSITIVE LOGITS

 corrected

0.99

correct

0.91

 Correct

0.90

 correction

0.88

Correct

0.87

 Correction

0.84

 correcting

0.84

Correction

0.82

 corrections

0.76

 spelling

0.75

Activations Density 0.014%

and highlight corrections or amendments in a text

instances of the word "correction" and its variations

No Comments

No Known Activations