INDEX

Explanations

phrases related to actions being taken or completed

oai_token-act-pair · gpt-3.5-turbo

the conjunction "and" 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

hov

-0.75

antz

-0.66

anu

-0.64

ci

-0.64

wrap

-0.64

nos

-0.63

Morning

-0.63

oids

-0.63

---------

-0.63

rain

-0.62

POSITIVE LOGITS

 therefore

1.10

 hence

1.09

 consequently

1.04

 thus

0.98

rogen

0.94

 thereby

0.94

rogens

0.94

 then

0.91

 redistributed

0.88

 secondly

0.85

Activations Density 0.312%

phrases related to actions being taken or completed

the conjunction "and" in various contexts

No Comments

No Known Activations