INDEX

Explanations

terms related to mental health, specifically the concept of lunacy

oai_token-act-pair · gpt-3.5-turbo

terms associated with "lunacy" and extreme political views

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

UV

-0.81

LECT

-0.77

ORN

-0.76

âĢ¢âĢ¢

-0.73

Mich

-0.72

EVA

-0.72

eff

-0.70

Ô

-0.70

 Norn

-0.70

Micro

-0.70

POSITIVE LOGITS

lun

1.35

acies

1.18

acy

0.93

atics

0.91

atic

0.86

gement

0.83

wich

0.83

atical

0.83

udicrous

0.81

acists

0.79

Activations Density 0.004%

terms related to mental health, specifically the concept of lunacy

terms associated with "lunacy" and extreme political views

No Comments

No Known Activations