INDEX

Explanations

numeric values

oai_token-act-pair · gpt-3.5-turbo

specific terms related to academic or professional contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

assies

-0.87

 Jagu

-0.73

bledon

-0.69

Ire

-0.69

DAQ

-0.67

drops

-0.64

 Oval

-0.64

elines

-0.64

 cabinets

-0.63

 Seym

-0.63

POSITIVE LOGITS

 Â·

0.80

 transcript

0.70

 Member

0.66

igmatic

0.64

 variant

0.63

 addict

0.62

)))

0.62

 Registered

0.61

 illustration

0.60

])

0.59

Activations Density 0.162%

numeric values

specific terms related to academic or professional contexts

No Comments

No Known Activations