INDEX

Explanations

news headlines involving police and law enforcement

oai_token-act-pair · gpt-3.5-turbo

terms related to police and incidents involving individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

lihood

-0.80

 Palest

-0.70

ctory

-0.67

ternity

-0.67

yss

-0.65

 Marketable

-0.65

agy

-0.63

Peb

-0.60

 Kinnikuman

-0.60

 Cheong

-0.59

POSITIVE LOGITS

itent

0.71

 district

0.66

 trooper

0.64

istrates

0.62

warts

0.61

 examiner

0.61

lance

0.61

 practitioner

0.60

 caucuses

0.60

itol

0.60

Activations Density 0.318%

news headlines involving police and law enforcement

terms related to police and incidents involving individuals

No Comments

No Known Activations