INDEX

Explanations

names of people or entities

oai_token-act-pair · gpt-3.5-turbo

proper nouns, specifically names and titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 proportion

-0.64

ãĥģ

-0.60

 FedEx

-0.60

 Thief

-0.60

 Costco

-0.58

 cartel

-0.56

 quadru

-0.56

AAA

-0.56

CTR

-0.56

Tup

-0.56

POSITIVE LOGITS

enegger

1.08

ricks

0.90

jen

0.87

yk

0.83

kson

0.83

enson

0.80

rov

0.75

esson

0.73

 recalled

0.73

sung

0.72

Activations Density 0.601%

names of people or entities

proper nouns, specifically names and titles

No Comments

No Known Activations