INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo

mentions of "Gra" or variations thereof in context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ional

-0.78

CPC

-0.77

 Dragonbound

-0.68

 Relations

-0.63

 Pixie

-0.63

Demand

-0.62

nikov

-0.61

 Nurs

-0.59

 Legion

-0.58

 Lear

-0.58

POSITIVE LOGITS

ffiti

1.55

veland

1.24

fters

1.22

veyard

1.20

fter

1.18

vel

1.17

ven

1.13

eme

1.13

ppa

1.10

ffen

1.07

Activations Density 0.026%

proper nouns

mentions of "Gra" or variations thereof in context

No Comments

No Known Activations