INDEX

Explanations

terms related to global scale, reach, or impact

oai_token-act-pair · gpt-3.5-turbo

key terms related to global systems and issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

tein

-0.60

formed

-0.58

66666666

-0.58

ilk

-0.57

 Melody

-0.57

 Wildcats

-0.57

 Palest

-0.56

bil

-0.56

bye

-0.56

alion

-0.56

POSITIVE LOGITS

ãĤ¼ãĤ¦ãĤ¹

0.91

rane

0.73

isd

0.73

abad

0.70

inois

0.69

uras

0.68

warts

0.68

 envoy

0.67

xia

0.67

warming

0.67

Activations Density 0.169%

terms related to global scale, reach, or impact

key terms related to global systems and issues

No Comments

No Known Activations