INDEX

Explanations

phrases prompting further exploration, investigation, or discussion

oai_token-act-pair · gpt-3.5-turbo

phrases that prompt examination or exploration of a topic

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

ELD

-0.87

Downloadha

-0.74

éĹ

-0.74

à¨

-0.72

CVE

-0.69

Unknown

-0.68

Course

-0.68

BuyableInstoreAndOnline

-0.65

ca

-0.65

auld

-0.65

POSITIVE LOGITS

amaz

0.81

querade

0.76

 ourselves

0.75

rg

0.71

acters

0.67

 reiterate

0.66

 clarify

0.64

00200000

0.64

dule

0.64

 illustrate

0.63

Activations Density 0.083%

phrases prompting further exploration, investigation, or discussion

phrases that prompt examination or exploration of a topic

No Comments

No Known Activations