INDEX

Explanations

references to a specific person named "Hub" with varying levels of importance or involvement

oai_token-act-pair · gpt-3.5-turbo

references to a specific entity or brand associated with technology or robotics

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

anguage

-0.92

Downloadha

-0.88

terday

-0.79

esthetic

-0.72

sembly

-0.69

vironment

-0.69

chnology

-0.68

 surviving

-0.68

ãĥ´

-0.68

 FORM

-0.68

POSITIVE LOGITS

bub

1.15

lot

1.12

hub

1.10

Hub

1.03

Hub

1.02

ris

0.99

master

0.97

bies

0.96

rod

0.95

bard

0.95

Activations Density 0.017%

references to a specific person named "Hub" with varying levels of importance or involvement

references to a specific entity or brand associated with technology or robotics

No Comments

No Known Activations