INDEX

Explanations

proper nouns related to people's names

oai_token-act-pair · gpt-3.5-turbo

references to the name "Howard" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

yrim

-0.77

asus

-0.74

erala

-0.71

uchin

-0.69

ffic

-0.66

eds

-0.66

ccording

-0.66

 circulate

-0.65

lopp

-0.65

undai

-0.64

POSITIVE LOGITS

 Howard

1.11

 Stern

1.06

 Ellis

0.91

Howard

0.86

 Schultz

0.81

 Buffett

0.78

 Dean

0.78

McN

0.78

 Faul

0.77

 Hughes

0.77

Activations Density 0.005%

proper nouns related to people's names

references to the name "Howard" in various contexts

No Comments

No Known Activations