INDEX

Explanations

words related to unrestrained or uncontrolled actions or concepts

oai_token-act-pair · gpt-3.5-turbo

words related to lack of restrictions or supervision

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

udeb

-0.83

bourg

-0.79

enegger

-0.76

ioxide

-0.73

utra

-0.72

anguage

-0.70

Clar

-0.69

oqu

-0.69

effective

-0.68

adder

-0.67

POSITIVE LOGITS

ly

0.77

ness

0.70

 pregnancies

0.68

adoes

0.67

 Territories

0.65

 territories

0.64

 territory

0.63

 baggage

0.61

uality

0.61

 curiosity

0.59

Activations Density 0.108%

words related to unrestrained or uncontrolled actions or concepts

words related to lack of restrictions or supervision

No Comments

No Known Activations