INDEX

Explanations

words related to holes or cutting an opening in something

oai_token-act-pair · gpt-3.5-turbo

references to physical holes or openings in various contexts

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

soDeliveryDate

-0.83

 Dispatch

-0.74

astical

-0.72

arcity

-0.71

DEM

-0.71

Online

-0.70

icip

-0.70

arming

-0.69

agy

-0.69

agger

-0.68

POSITIVE LOGITS

 protr

1.18

 holes

1.13

 cavity

1.01

 diameter

0.97

 hole

0.97

 membranes

0.97

 casing

0.97

 thicker

0.91

 insulation

0.91

 seams

0.89

Activations Density 0.369%

words related to holes or cutting an opening in something

references to physical holes or openings in various contexts

No Comments

No Known Activations