INDEX

Explanations

mentions of locations, especially those starting with the letter "B"

oai_token-act-pair · gpt-3.5-turbo

capitalized entities, specifically proper nouns

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Hilbert

-0.61

 dere

-0.59

 dues

-0.59

Bet

-0.59

 hear

-0.58

 Revel

-0.57

due

-0.57

 acne

-0.56

 posit

-0.56

 Mole

-0.56

POSITIVE LOGITS

amboo

1.33

ishops

1.25

izarre

1.20

ibli

1.19

isexual

1.18

antam

1.18

odies

1.18

ourgeois

1.17

asket

1.16

ikini

1.14

Activations Density 0.042%

mentions of locations, especially those starting with the letter "B"

capitalized entities, specifically proper nouns

No Comments

No Known Activations