INDEX

Explanations

words related to mystery or secrecy

oai_token-act-pair · gpt-3.5-turbo

references to the concept of mystery or things that are described as mysterious

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

hers

-0.83

©¶æ

-0.79

arf

-0.79

berman

-0.76

lag

-0.75

midt

-0.73

dh

-0.70

win

-0.69

rikes

-0.68

bers

-0.67

POSITIVE LOGITS

 mysterious

0.97

 unexplained

0.88

 unsolved

0.87

 mysteriously

0.81

ly

0.79

 suspic

0.79

 mystery

0.78

 enigmatic

0.78

 disappearance

0.76

worldly

0.75

Activations Density 0.011%

words related to mystery or secrecy

references to the concept of mystery or things that are described as mysterious

No Comments

No Known Activations