INDEX

Explanations

references to large urban areas or cities

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "meg" 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 @ 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

 Sense

-0.65

 cover

-0.62

 false

-0.61

 Volks

-0.61

 Mercury

-0.60

++++++++++++++++

-0.59

ãģį

-0.59

 condition

-0.58

 Taste

-0.57

 Mesa

-0.57

POSITIVE LOGITS

abytes

1.44

awatts

1.43

abyte

1.40

apixel

1.37

acity

1.23

aton

1.20

apolis

1.18

aph

1.17

amorph

1.15

atis

1.15

Activations Density 0.023%

references to large urban areas or cities

occurrences of the word "meg" in various contexts

No Comments

No Known Activations