INDEX

Explanations

words related to companies or organizations

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 TAMADRA

-0.86

oldemort

-0.73

bes

-0.73

ixel

-0.73

swick

-0.69

bos

-0.67

 Required

-0.67

>[

-0.66

figure

-0.66

Cola

-0.66

POSITIVE LOGITS

 preservation

0.86

 related

0.84

 logistical

0.76

 aviation

0.74

aer

0.74

 conservation

0.74

 Related

0.73

 entertainment

0.73

 literacy

0.72

 maintenance

0.72

Activations Density 0.203%

words related to companies or organizations

No Comments

No Known Activations