INDEX

Explanations

words related to coding languages and programming concepts, such as functions and accelerated courses

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

inventoryQuantity

-0.87

estate

-0.85

soType

-0.79

agate

-0.74

advertising

-0.72

ghazi

-0.72

ãĥ´ãĤ¡

-0.72

poll

-0.71

 billboard

-0.71

aretz

-0.70

POSITIVE LOGITS

 Instructor

1.92

 instructors

1.85

 curriculum

1.81

 instructor

1.76

Course

1.67

 Teaching

1.67

 curric

1.66

 Instruct

1.66

 learners

1.66

 lectures

1.64

Activations Density 0.721%

words related to coding languages and programming concepts, such as functions and accelerated courses

No Comments

No Known Activations