INDEX

Explanations

phrases related to presentations, discussions, and lectures

oai_token-act-pair · gpt-3.5-turbo

instances of the word "the"

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

Iterator

-0.71

âĢł

-0.71

 depended

-0.66

strap

-0.65

--+

-0.64

Layer

-0.64

 debuted

-0.62

 cheered

-0.62

ESPN

-0.61

Chart

-0.61

POSITIVE LOGITS

 possibility

1.14

 impending

1.08

 latest

1.03

 dangers

1.03

 extent

1.02

 specifics

1.01

 latter

1.01

 upcoming

1.00

 workings

0.99

 importance

0.98

Activations Density 0.504%

phrases related to presentations, discussions, and lectures

instances of the word "the"

No Comments

No Known Activations