INDEX

Explanations

phrases indicating coordination or connection between different entities

oai_token-act-pair · gpt-3.5-turbo

phrases that suggest relationships, collaborations, or combinations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

quickShipAvailable

-0.55

enthal

-0.53

ampa

-0.51

°

-0.51

andum

-0.48

åī

-0.48

zeb

-0.47

agara

-0.47

ints

-0.47

itary

-0.46

POSITIVE LOGITS

 with

1.90

with

1.70

 WITH

1.60

 With

1.53

 withd

1.47

With

1.41

wi

0.99

WithNo

0.97

wit

0.93

amorph

0.68

Activations Density 2.310%

phrases indicating coordination or connection between different entities

phrases that suggest relationships, collaborations, or combinations

No Comments

No Known Activations