INDEX

Explanations

words related to ballot propositions or laws

oai_token-act-pair · gpt-3.5-turbo

references to specific ballot propositions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

soever

-0.83

esome

-0.71

Interstitial

-0.69

 Woods

-0.67

 GOODMAN

-0.67

keeper

-0.66

 Ashe

-0.64

 Carib

-0.64

 fingertips

-0.64

 Sins

-0.64

POSITIVE LOGITS

onents

1.33

onent

1.30

ulsion

1.24

aganda

1.23

osed

1.23

ensity

1.20

osition

1.19

ulsive

1.06

hetical

1.02

hetic

1.02

Activations Density 0.018%

words related to ballot propositions or laws

references to specific ballot propositions

No Comments

No Known Activations