INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo

proper names or identifiers, especially related to individuals and entities

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

¶ħ

-0.52

DOI

-0.52

edition

-0.47

/_

-0.45

-0.44

į

-0.43

idth

-0.42

ä»£

-0.42

 FedEx

-0.41

************

-0.41

POSITIVE LOGITS

espie

0.58

raltar

0.49

lings

0.49

yang

0.48

oak

0.47

zon

0.47

ython

0.44

 detractors

0.44

enson

0.44

eton

0.43

Activations Density 0.641%

proper nouns

proper names or identifiers, especially related to individuals and entities

No Comments

No Known Activations