INDEX

Explanations

prefixes "ex-" followed by a word

oai_token-act-pair · gpt-3.5-turbo

the term "ex" as a prefix indicating former status or past relationships

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

 GOODMAN

-0.83

ãĤ¤ãĥĪ

-0.82

 Sabha

-0.82

jriwal

-0.76

WAYS

-0.73

GoldMagikarp

-0.71

 Flavoring

-0.69

arity

-0.68

WARE

-0.65

hower

-0.65

POSITIVE LOGITS

clamation

1.20

uber

1.08

ogenous

1.05

orbit

1.04

terior

1.03

oplan

1.03

clud

1.03

uding

0.98

communication

0.94

uded

0.93

Activations Density 0.010%

prefixes "ex-" followed by a word

the term "ex" as a prefix indicating former status or past relationships

No Comments

No Known Activations