INDEX

Explanations

proper nouns related to people

oai_token-act-pair · gpt-3.5-turbo

occurrences of names and titles of people

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

sylv

-0.76

 sarc

-0.71

acebook

-0.70

 sugg

-0.68

 fict

-0.67

SOURCE

-0.67

idon

-0.66

avorite

-0.66

inarily

-0.64

 clos

-0.64

POSITIVE LOGITS

Daw

0.94

 Sabha

0.91

 Gandhi

0.83

Sek

0.73

uri

0.72

 Khalid

0.71

Yak

0.69

 Chow

0.69

 Krish

0.68

 Sharma

0.67

Activations Density 0.094%

proper nouns related to people

occurrences of names and titles of people

No Comments

No Known Activations