INDEX

Explanations

statements or descriptions related to criticism or controversy

oai_token-act-pair · gpt-3.5-turbo

elements indicating conflict or tension within narratives

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

ITNESS

-0.98

iterranean

-0.56

itable

-0.55

version

-0.55

ItemThumbnailImage

-0.55

PN

-0.55

Recommended

-0.54

 Might

-0.53

 millenn

-0.52

æĢ

-0.52

POSITIVE LOGITS

 onto

1.43

 into

1.41

 away

1.34

 goodbye

1.32

 overboard

1.29

 aside

1.23

 apart

1.20

 down

1.18

off

1.15

 INTO

1.13

Activations Density 0.660%

statements or descriptions related to criticism or controversy

elements indicating conflict or tension within narratives

No Comments

No Known Activations