INDEX

Explanations

names of individuals

oai_token-act-pair · gpt-3.5-turbo

proper nouns of individuals or significant entities

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

 depreciation

-0.78

 concessions

-0.74

 subsistence

-0.73

 ACTION

-0.72

 dracon

-0.71

 Skydragon

-0.70

 Cerberus

-0.69

è¦ļéĨĴ

-0.68

 punishable

-0.67

 Pastebin

-0.67

POSITIVE LOGITS

inda

1.37

ona

1.31

ilyn

1.26

isha

1.21

issa

1.19

aleigh

1.17

annah

1.16

iona

1.16

lene

1.15

endra

1.13

Activations Density 0.352%

names of individuals

proper nouns of individuals or significant entities

No Comments

No Known Activations