INDEX

Explanations

mentions of specific names, likely related to incidents or people

oai_token-act-pair · gpt-3.5-turbo

specific mentions of the name "Columbine" and related entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

bourg

-0.78

ãģ¦

-0.73

achus

-0.64

erous

-0.63

boards

-0.61

charge

-0.61

bilt

-0.61

ciation

-0.61

 tiss

-0.60

Sing

-0.60

POSITIVE LOGITS

 adren

0.79

iHUD

0.75

gio

0.73

icultural

0.72

Eva

0.72

 Reloaded

0.71

 Varg

0.70

iculture

0.69

amins

0.68

Val

0.68

Activations Density 0.013%

mentions of specific names, likely related to incidents or people

specific mentions of the name "Columbine" and related entities

No Comments

No Known Activations