INDEX

Explanations

phrases related to permissions, agreements, and command-like statements

oai_token-act-pair · gpt-3.5-turbo

connective words and phrases that suggest agreements or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 Doodle

-0.74

 âĢİ

-0.71

 Companion

-0.65

inence

-0.64

ounter

-0.63

å£«

-0.62

"],

-0.62

cock

-0.62

pher

-0.61

isphere

-0.61

POSITIVE LOGITS

 rightly

0.76

 unsuccessfully

0.75

 Bezos

0.68

 Prosecut

0.66

 unsurprisingly

0.64

acter

0.63

bid

0.63

 indeed

0.62

 Miliband

0.62

 goalt

0.61

Activations Density 0.386%

phrases related to permissions, agreements, and command-like statements

connective words and phrases that suggest agreements or conditions

No Comments

No Known Activations