INDEX

Explanations

references to locations, especially in or related to New York

oai_token-act-pair · gpt-3.5-turbo

the letter "Y" 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 @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãĥ¼ãĥĨãĤ£

-0.67

ãĥ¯

-0.65

ĸļ

-0.65

ãĥķãĤ©

-0.65

raints

-0.64

 facilitated

-0.64

 foil

-0.64

inelli

-0.62

cientious

-0.62

idity

-0.62

POSITIVE LOGITS

STEM

1.01

ORK

1.01

ARD

0.97

BI

0.94

RS

0.93

ANK

0.93

ANG

0.92

ERE

0.91

0.90

UGE

0.90

Activations Density 0.016%

references to locations, especially in or related to New York

the letter "Y" in various contexts

No Comments

No Known Activations