INDEX

Explanations

words related to legality and illegality, such as "illegal," "legal," and "crime."

oai_token-act-pair · gpt-3.5-turbo

terms related to legality and illegal activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ochond

-0.76

idges

-0.71

pend

-0.70

ellect

-0.68

borough

-0.65

ras

-0.65

hetic

-0.64

ortun

-0.64

 dilig

-0.64

ases

-0.63

POSITIVE LOGITS

 punishable

1.03

 federally

0.87

 under

0.81

 unless

0.79

 violation

0.78

 legally

0.78

 except

0.77

 everywhere

0.76

 nationwide

0.75

 internationally

0.75

Activations Density 0.117%

words related to legality and illegality, such as "illegal," "legal," and "crime."

terms related to legality and illegal activities

No Comments

No Known Activations