INDEX

Explanations

numerical values or measurements with the word 'single'

oai_token-act-pair · gpt-3.5-turbo

references to "single-payer" health care systems

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

Downloadha

-0.72

raints

-0.71

æ©Ł

-0.68

hips

-0.67

heny

-0.63

clinton

-0.63

 Guerrero

-0.62

 Lauder

-0.61

ravings

-0.61

utterstock

-0.60

POSITIVE LOGITS

handedly

1.56

ton

1.20

player

1.03

 digits

1.02

 malt

1.01

 digit

1.00

 minded

0.97

minded

0.96

 handed

0.94

 sided

0.92

Activations Density 0.043%

numerical values or measurements with the word 'single'

references to "single-payer" health care systems

No Comments

No Known Activations