INDEX

Explanations

locations

oai_token-act-pair · gpt-3.5-turbo

instances of the word "in" used throughout various contexts or locations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

hemy

-0.71

 inval

-0.65

 announ

-0.65

 guiActiveUn

-0.65

henko

-0.63

ogether

-0.63

hirt

-0.63

killed

-0.62

ratulations

-0.62

ictive

-0.61

POSITIVE LOGITS

 lieu

1.28

 conjunction

1.23

 accordance

1.08

 vitro

1.01

clus

1.00

patient

0.94

 Switzerland

0.92

 downtown

0.92

efficiency

0.90

 order

0.90

Activations Density 0.457%

locations

instances of the word "in" used throughout various contexts or locations

No Comments

No Known Activations