INDEX

Explanations

references to behaviors related to rules, ethics, and regulations

oai_token-act-pair · gpt-3.5-turbo

instances of the word "conduct" and related forms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

cube

-0.73

corn

-0.64

ched

-0.62

fork

-0.61

 Daddy

-0.60

ches

-0.59

added

-0.59

loo

-0.58

Lenin

-0.58

Todd

-0.57

POSITIVE LOGITS

ors

1.18

ivity

1.13

ivities

1.02

atform

0.90

ional

0.90

ORS

0.86

ively

0.85

uations

0.84

ored

0.84

ives

0.84

Activations Density 0.043%

references to behaviors related to rules, ethics, and regulations

instances of the word "conduct" and related forms

No Comments

No Known Activations