INDEX

Explanations

technical terms and product names in a variety of languages

oai_token-act-pair · gpt-3.5-turbo

proper nouns and names in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

IM

-0.60

 DIRECT

-0.59

COM

-0.58

 Chambers

-0.56

IMP

-0.56

REC

-0.56

 WITHOUT

-0.55

 MILL

-0.55

BIG

-0.54

gra

-0.54

POSITIVE LOGITS

assic

0.96

escription

0.95

enium

0.84

oan

0.84

emonic

0.82

illac

0.82

aq

0.81

eto

0.80

aceae

0.79

Activations Density 0.334%

technical terms and product names in a variety of languages

proper nouns and names in the text

No Comments

No Known Activations