INDEX

Explanations

evidence or items discovered by authorities during investigations

oai_token-act-pair · gpt-3.5-turbo

instances of the word "found" associated with events or discoveries

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

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

externalActionCode

-0.75

 guiActiveUn

-0.74

é¾įå¥ĳå£«

-0.73

adan

-0.72

 privile

-0.70

IQ

-0.70

panic

-0.69

ignore

-0.68

selling

-0.67

 captcha

-0.65

POSITIVE LOGITS

ered

1.06

 traces

1.00

 evidence

0.93

 remnants

0.85

 nothing

0.85

 pockets

0.81

 plenty

0.80

 footprints

0.77

 ample

0.77

 dozens

0.75

Activations Density 0.076%

evidence or items discovered by authorities during investigations

instances of the word "found" associated with events or discoveries

No Comments

No Known Activations