INDEX

Explanations

phrases describing a process of creating something or utilizing specific tools or techniques

oai_token-act-pair · gpt-3.5-turbo

the preposition "to" indicating intent or direction

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

TION

-0.72

CI

-0.71

listed

-0.68

angered

-0.63

lich

-0.62

going

-0.62

nin

-0.62

 Hours

-0.60

 regrets

-0.59

soDeliveryDate

-0.58

POSITIVE LOGITS

 maximize

1.42

 create

1.35

 enhance

1.35

 achieve

1.35

 ensure

1.32

 determine

1.28

 optimize

1.27

 simplify

1.23

 facilitate

1.18

 justify

1.18

Activations Density 0.278%

phrases describing a process of creating something or utilizing specific tools or techniques

the preposition "to" indicating intent or direction

No Comments

No Known Activations