INDEX

Explanations

concepts related to authority or extremity

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "absolute" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

NetMessage

-0.91

nan

-0.85

enegger

-0.85

anners

-0.84

bucks

-0.83

actionDate

-0.82

ACP

-0.80

boards

-0.80

board

-0.78

engers

-0.78

POSITIVE LOGITS

 monarchy

1.08

 monarch

0.82

 freedom

0.80

 majority

0.79

 necessity

0.79

 positioning

0.79

 importance

0.78

 equival

0.77

 domination

0.77

 separation

0.76

Activations Density 0.011%

concepts related to authority or extremity

occurrences of the word "absolute" in various contexts

No Comments

No Known Activations