INDEX

Explanations

words related to habits and patterns

oai_token-act-pair · gpt-3.5-turbo

references to various habits and societal norms

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

MER

-0.80

bilt

-0.68

moon

-0.64

ded

-0.64

ESA

-0.63

plane

-0.63

NX

-0.62

onement

-0.62

jet

-0.62

×©

-0.61

POSITIVE LOGITS

omething

1.24

ystem

1.11

uggest

1.10

cape

1.09

poons

1.06

hops

1.05

etter

1.03

hips

1.02

ettings

1.01

mith

0.99

Activations Density 0.173%

words related to habits and patterns

references to various habits and societal norms

No Comments

No Known Activations