INDEX

Explanations

instances where someone or something is held responsible for certain actions or situations

oai_token-act-pair · gpt-3.5-turbo

phrases related to accountability and responsibility

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

Sport

-0.82

quickShipAvailable

-0.79

 guiActiveUnfocused

-0.76

edin

-0.72

grand

-0.72

æĺ

-0.72

isSpecialOrderable

-0.71

edu

-0.71

IQ

-0.70

DragonMagazine

-0.70

POSITIVE LOGITS

 overseeing

1.25

 maintaining

1.20

 ensuring

1.17

 upholding

1.10

 safegu

1.07

 protecting

1.05

 creating

1.04

 regulating

1.01

 keeping

1.01

 implementing

1.00

Activations Density 0.082%

instances where someone or something is held responsible for certain actions or situations

phrases related to accountability and responsibility

No Comments

No Known Activations