INDEX

Explanations

words related to safety or protection

oai_token-act-pair · gpt-3.5-turbo

terms related to safety and protection in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

awks

-0.71

awk

-0.70

ancies

-0.68

yss

-0.64

 gust

-0.64

ENG

-0.63

ipop

-0.62

awar

-0.62

ikarp

-0.62

plet

-0.61

POSITIVE LOGITS

ablishment

0.71

 fodder

0.71

bridge

0.71

DragonMagazine

0.69

 scapego

0.68

 discipl

0.67

SAN

0.66

cially

0.66

ãĥķãĤ¡

0.65

PsyNetMessage

0.65

Activations Density 0.344%

words related to safety or protection

terms related to safety and protection in various contexts

No Comments

No Known Activations