INDEX

Explanations

interactions involving communication or authority figures

oai_token-act-pair · gpt-3.5-turbo

declarative statements indicating refusal or denial

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Inher

-0.86

 ageing

-0.80

 renaissance

-0.72

 celebrates

-0.71

 extinct

-0.71

 relegation

-0.71

 nostalgic

-0.68

 evolutionary

-0.67

 vintage

-0.67

obyl

-0.66

POSITIVE LOGITS

 subpoena

1.15

 subpoen

1.11

Request

1.04

 requesting

1.01

 instructed

1.01

 request

1.01

 confidentiality

1.01

 politely

0.98

fax

0.96

 requests

0.89

Activations Density 1.283%

interactions involving communication or authority figures

declarative statements indicating refusal or denial

No Comments

No Known Activations