INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo

names and mentions of notable individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

lords

-0.90

meal

-0.78

 Hussein

-0.75

Demand

-0.71

manship

-0.70

lord

-0.68

 Declaration

-0.64

 Clockwork

-0.63

 Luxem

-0.63

mann

-0.62

POSITIVE LOGITS

rique

0.92

rics

0.89

tered

0.84

icol

0.83

rika

0.77

avid

0.76

icle

0.75

agraph

0.75

icles

0.74

ersen

0.74

Activations Density 0.015%

proper nouns

names and mentions of notable individuals

No Comments

No Known Activations