INDEX

Explanations

names of individuals

oai_token-act-pair · gpt-3.5-turbo

proper nouns, specifically names of people

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

cake

-0.65

OPS

-0.62

 Sussex

-0.61

 Legion

-0.60

 Cure

-0.60

ually

-0.60

 Blaze

-0.60

jack

-0.60

Word

-0.60

FW

-0.59

POSITIVE LOGITS

quist

1.19

gren

1.17

sky

1.05

kson

0.99

qv

0.90

enegger

0.90

hetti

0.90

chuk

0.89

afort

0.89

ramid

0.88

Activations Density 0.022%

names of individuals

proper nouns, specifically names of people

No Comments

No Known Activations