INDEX

Explanations

terms related to crime, law enforcement, investigations, and related activities

oai_token-act-pair · gpt-3.5-turbo

words related to crime and criminality

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

 Lisbon

-0.70

esc

-0.66

 scratch

-0.66

CAR

-0.64

ETF

-0.63

Constructed

-0.63

 Mandela

-0.61

 Dangerous

-0.61

 Mississ

-0.61

 Senegal

-0.61

POSITIVE LOGITS

nant

1.02

onial

0.99

rim

0.97

inatory

0.96

inals

0.93

inant

0.89

inally

0.88

inating

0.88

etric

0.86

ming

0.86

Activations Density 0.018%

terms related to crime, law enforcement, investigations, and related activities

words related to crime and criminality

No Comments

No Known Activations