INDEX

Explanations

information related to percentages and quantities

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

epad

-0.64

 Multiple

-0.59

 Norn

-0.59

Iss

-0.58

 Increasing

-0.58

bart

-0.57

itely

-0.56

 Encyclopedia

-0.54

 Dictionary

-0.54

IENT

-0.53

POSITIVE LOGITS

 revolves

1.10

 resides

1.05

 consisted

1.02

 originate

1.01

 reside

0.99

 belonged

0.98

 stemmed

0.98

 derives

0.97

 consists

0.95

 stems

0.94

Activations Density 0.353%

information related to percentages and quantities

No Comments

No Known Activations