INDEX

Explanations

proper names or terms containing "Nak" or "Nad" in them

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals, particularly notable names

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

acebook

-0.77

BuyableInstoreAndOnline

-0.71

 ILCS

-0.68

DVD

-0.66

 brim

-0.65

 seas

-0.63

ICS

-0.63

 Chill

-0.63

Pyr

-0.62

 sexes

-0.61

POSITIVE LOGITS

amoto

1.13

amura

1.05

otes

0.97

ay

0.84

iri

0.83

hod

0.82

Nak

0.82

enta

0.82

gement

0.81

gments

0.81

Activations Density 0.042%

proper names or terms containing "Nak" or "Nad" in them

references to specific individuals, particularly notable names

No Comments

No Known Activations