INDEX

Explanations

mentions of the Medicare program or related terms

oai_token-act-pair · gpt-3.5-turbo

references to Medicare

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

amy

-0.70

imov

-0.69

semble

-0.67

variable

-0.65

quart

-0.63

amps

-0.63

 Dresden

-0.62

=-=-=-=-=-=-=-=-

-0.61

rings

-0.60

ORE

-0.60

POSITIVE LOGITS

 Medicare

0.93

 reimbursement

0.88

 beneficiaries

0.87

 entitle

0.86

 Advantage

0.85

amacare

0.84

 concess

0.78

 fare

0.77

 reimburse

0.76

 Benefits

0.75

Activations Density 0.006%

mentions of the Medicare program or related terms

references to Medicare

No Comments

No Known Activations