INDEX

Explanations

references to bilateral interactions or agreements

oai_token-act-pair · gpt-3.5-turbo

mentions of opposing factions or parties in a conflict

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

RON

-0.67

 dehyd

-0.67

 curfew

-0.64

ardy

-0.63

nutrition

-0.61

AUT

-0.61

nce

-0.61

interrupted

-0.60

 Raven

-0.60

Dub

-0.60

POSITIVE LOGITS

pace

0.93

 sides

0.91

 thereof

0.83

side

0.82

cape

0.80

 concede

0.77

lip

0.77

addle

0.77

orts

0.73

hops

0.72

Activations Density 0.018%

references to bilateral interactions or agreements

mentions of opposing factions or parties in a conflict

No Comments

No Known Activations