INDEX

Explanations

references to ideals or idealized concepts

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "ideal" 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 @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

bane

-0.74

cloth

-0.68

SEE

-0.65

Lago

-0.63

cano

-0.63

rup

-0.61

NetMessage

-0.61

words

-0.61

âĵĺ

-0.60

leness

-0.59

POSITIVE LOGITS

istic

1.65

istically

1.46

ized

1.37

ised

1.32

ization

1.18

isation

1.13

izations

1.11

izable

1.09

isable

1.08

izing

1.08

Activations Density 0.036%

references to ideals or idealized concepts

references to the concept of "ideal" in various contexts

No Comments

No Known Activations