INDEX

Explanations

obscure or hidden information

oai_token-act-pair · gpt-3.5-turbo

references to the concept of obscurity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

hner

-0.79

Demand

-0.76

ioxide

-0.71

utherford

-0.70

avers

-0.68

hao

-0.66

onz

-0.64

uana

-0.64

=-=-=-=-=-=-=-=-

-0.64

eng

-0.64

POSITIVE LOGITS

 obscure

1.20

 obscurity

1.11

uously

0.84

ously

0.72

 niche

0.72

 forgotten

0.71

ulously

0.71

 trivia

0.70

ingly

0.70

neb

0.69

Activations Density 0.007%

obscure or hidden information

references to the concept of obscurity

No Comments

No Known Activations