INDEX

Explanations

mentions of the term "Medicare"

oai_token-act-pair · gpt-3.5-turbo

references to the Medicare program

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

kan

-0.65

jin

-0.64

NK

-0.62

 blaster

-0.62

uin

-0.61

oun

-0.60

oon

-0.60

Jer

-0.59

Storm

-0.59

 storm

-0.58

POSITIVE LOGITS

 Medicare

3.73

 Medicaid

2.21

Medic

1.96

 ObamaCare

1.62

 Obamacare

1.62

HHS

1.44

ACA

1.40

 Healthcare

1.35

payer

1.35

NHS

1.33

Activations Density 0.011%

mentions of the term "Medicare"

references to the Medicare program

No Comments

No Known Activations