INDEX

Explanations

questions being asked by interviewers to guests or public figures

oai_token-act-pair · gpt-3.5-turbo

instances of questioning or inquiry in discussions or interviews

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

brance

-0.74

Torrent

-0.74

 Maiden

-0.71

soDeliveryDate

-0.70

 execut

-0.68

 upstream

-0.68

 ILCS

-0.67

celona

-0.67

 Kirin

-0.65

 harvest

-0.65

POSITIVE LOGITS

 incred

0.96

 rhet

0.94

Asked

0.93

 Kimmel

0.92

 condesc

0.92

 filib

0.89

 moderator

0.88

question

0.87

 chuck

0.87

kward

0.87

Activations Density 0.539%

questions being asked by interviewers to guests or public figures

instances of questioning or inquiry in discussions or interviews

No Comments

No Known Activations