INDEX

Explanations

instances where credit is being given to someone or something

oai_token-act-pair · gpt-3.5-turbo

references to giving credit or acknowledgment to individuals or entities

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

 Lans

-0.83

dq

-0.67

Osw

-0.65

 Somers

-0.65

ften

-0.63

 intest

-0.63

 Weird

-0.62

improve

-0.62

vae

-0.61

 Dresden

-0.61

POSITIVE LOGITS

card

0.82

worthiness

0.82

 card

0.80

olor

0.80

 CARD

0.79

Credit

0.79

 credit

0.79

enza

0.78

itism

0.77

cards

0.74

Activations Density 0.018%

instances where credit is being given to someone or something

references to giving credit or acknowledgment to individuals or entities

No Comments

No Known Activations