INDEX

Explanations

proper nouns or specific phrases known by their unique terminology

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate a definition or explanation starting with the word "known as."

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

eor

-0.75

midt

-0.72

arnaev

-0.72

essim

-0.72

ORTS

-0.67

ubb

-0.67

olitan

-0.66

uliffe

-0.65

ebted

-0.65

ersen

-0.64

POSITIVE LOGITS

pires

0.76

 follows

0.71

COP

0.69

calling

0.69

 Fancy

0.65

phy

0.65

phe

0.65

pins

0.65

pas

0.62

Activations Density 0.048%

proper nouns or specific phrases known by their unique terminology

phrases that indicate a definition or explanation starting with the word "known as."

No Comments

No Known Activations