INDEX

Explanations

expletives and vulgar language

oai_token-act-pair · gpt-3.5-turbo

variations of profanity and explicit language

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Flavoring

-0.76

Level

-0.73

icles

-0.71

ntil

-0.68

Interstitial

-0.67

knit

-0.67

vale

-0.65

endant

-0.65

Footnote

-0.64

Serv

-0.64

POSITIVE LOGITS

 kidding

1.00

 stupid

0.85

 retarded

0.84

 idiot

0.80

 nuts

0.80

THING

0.79

 sucks

0.78

 bastard

0.77

 retard

0.77

 hell

0.77

Activations Density 0.063%

expletives and vulgar language

variations of profanity and explicit language

No Comments

No Known Activations