INDEX

Explanations

references to entities or terms related to the United States (U.S.)

oai_token-act-pair · gpt-3.5-turbo

references to the United States

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 bars

-0.66

 Discipline

-0.66

 rainbow

-0.61

 venue

-0.58

 Wicked

-0.57

 lockout

-0.56

 Clown

-0.56

 stiffness

-0.56

 bubble

-0.56

 Held

-0.55

POSITIVE LOGITS

topia

1.09

lyss

1.02

nexpected

1.01

zbek

0.99

llah

0.99

PDATED

0.99

seless

0.98

gly

0.96

rine

0.95

Conn

0.95

Activations Density 0.042%

references to entities or terms related to the United States (U.S.)

references to the United States

No Comments

No Known Activations