INDEX

Explanations

phrases related to being in charge or controlling something

oai_token-act-pair · gpt-3.5-turbo

references to groups, organizations, or entities that exert control or influence

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ayers

-0.83

peak

-0.73

ema

-0.73

 Chomsky

-0.72

oult

-0.71

orrow

-0.70

hesda

-0.70

lime

-0.69

iatus

-0.69

eeks

-0.65

POSITIVE LOGITS

 smoothly

0.99

 simulations

0.89

gam

0.87

err

0.86

 affairs

0.86

auntlet

0.86

 risk

0.84

 efficiently

0.77

 tests

0.76

ways

0.76

Activations Density 0.142%

phrases related to being in charge or controlling something

references to groups, organizations, or entities that exert control or influence

No Comments

No Known Activations