INDEX

Explanations

descriptions of services or products and related details such as country of origin and features

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate responsibility and understanding in service-oriented contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 Dancing

-0.68

Brit

-0.67

 Wade

-0.67

Surv

-0.67

Xan

-0.66

 triumphant

-0.66

 Amid

-0.65

 Gone

-0.64

Mom

-0.63

 Gong

-0.62

POSITIVE LOGITS

 certify

1.25

 indemn

1.23

 subcontract

1.23

 verify

1.20

 administer

1.19

 reimburse

1.18

 notify

1.18

 provide

1.17

 authorize

1.16

 invoice

1.15

Activations Density 0.538%

descriptions of services or products and related details such as country of origin and features

phrases that indicate responsibility and understanding in service-oriented contexts

No Comments

No Known Activations