INDEX

Explanations

documents, reports, and information sources referenced in sentences

oai_token-act-pair · gpt-3.5-turbo

references to various types of reports and documents 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 @ 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

abiding

-0.86

obyl

-0.81

btn

-0.77

quit

-0.75

ĸļ

-0.70

akeru

-0.68

successfully

-0.67

rontal

-0.65

ogging

-0.64

together

-0.63

POSITIVE LOGITS

 tracker

0.78

 insider

0.70

pedia

0.67

 datas

0.67

glers

0.65

 interviewer

0.65

Fib

0.64

 citation

0.63

 Fortune

0.62

 recru

0.62

Activations Density 0.277%

documents, reports, and information sources referenced in sentences

references to various types of reports and documents in the text

No Comments

No Known Activations