INDEX

Explanations

proper nouns related to military or navy entities

oai_token-act-pair · gpt-3.5-turbo

references to specific organizations or institutions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

olkien

-0.68

aeper

-0.67

Ara

-0.67

hue

-0.67

Wah

-0.66

ampoo

-0.65

Wem

-0.63

icago

-0.63

mble

-0.62

vertisements

-0.62

POSITIVE LOGITS

 Fired

0.69

created

0.68

create

0.65

partial

0.65

cedented

0.65

phrine

0.63

 Author

0.63

spons

0.62

asonable

0.61

rely

0.60

Activations Density 0.000%

proper nouns related to military or navy entities

references to specific organizations or institutions

No Comments

No Known Activations