INDEX

Explanations

words related to negative events or situations

oai_token-act-pair · gpt-3.5-turbo

negative sentiments or criticisms about various topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 Dickinson

-0.85

ulhu

-0.85

 EntityItem

-0.72

ħĭ

-0.69

 Norris

-0.66

AVG

-0.65

HCR

-0.65

 Compass

-0.65

 Vessel

-0.64

DragonMagazine

-0.63

POSITIVE LOGITS

based

1.29

sized

1.28

level

1.12

themed

1.06

related

1.06

induced

1.04

style

1.04

shaped

1.03

derived

1.02

powered

1.01

Activations Density 0.547%

words related to negative events or situations

negative sentiments or criticisms about various topics

No Comments

No Known Activations