INDEX

Explanations

references to incidents and activities involving law enforcement and crime

oai_token-act-pair · gpt-3.5-turbo

phrases related to accidents and damages

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

Ranked

-0.89

inction

-0.84

inav

-0.80

iliation

-0.80

thood

-0.79

è¦ļéĨĴ

-0.75

 Countries

-0.74

rowth

-0.74

meric

-0.74

ouble

-0.73

POSITIVE LOGITS

 homeowner

1.28

 cops

1.23

 shoppers

1.12

 investigators

1.11

 Craigslist

1.10

 firefighters

1.10

 thieves

1.09

 technicians

1.09

 neighbors

1.09

TMZ

1.08

Activations Density 0.910%

references to incidents and activities involving law enforcement and crime

phrases related to accidents and damages

No Comments

No Known Activations