INDEX

Explanations

coding-related terms and possibly error messages

oai_token-act-pair · gpt-3.5-turbo

structured data, particularly in the form of error messages and code snippets

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

 endings

-0.67

 impunity

-0.66

leans

-0.66

 abduct

-0.66

ĺħ

-0.65

rats

-0.65

 conclud

-0.63

 confir

-0.63

ongyang

-0.62

 subsequ

-0.62

POSITIVE LOGITS

":{"

1.10

quickShipAvailable

1.00

":["

0.94

":[{"

0.89

tnc

0.88

ItemImage

0.88

isSpecialOrderable

0.83

ãĥĺãĥ©

0.83

soType

0.79

inventoryQuantity

0.79

Activations Density 0.057%

coding-related terms and possibly error messages

structured data, particularly in the form of error messages and code snippets

No Comments

No Known Activations