INDEX

Explanations

programming-related keywords and actions

oai_token-act-pair · gpt-3.5-turbo

instances of the pronoun "I" indicating personal involvement or action

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

IFrame

Link

Negative Logits

 Gale

-0.65

 independence

-0.63

 careers

-0.63

 Jarrett

-0.63

 Rockefeller

-0.61

tnc

-0.60

 consequences

-0.59

Yon

-0.59

 souls

-0.58

 Newt

-0.58

POSITIVE LOGITS

'm

1.34

've

1.17

'll

1.00

WI

0.99

EEE

0.98

 intend

0.98

 personally

0.96

am

0.95

 prefer

0.92

WB

0.92

Activations Density 0.265%

programming-related keywords and actions

instances of the pronoun "I" indicating personal involvement or action

No Known Activations