INDEX

Explanations

words related to specific names or terms such as "Middlesbrough," "numeracy," and "Northeastern."

oai_token-act-pair · gpt-3.5-turbo

references to specific places and names, particularly in a statistical or comparative context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Benson

-0.68

 Mehran

-0.68

 Hands

-0.68

issance

-0.67

ACP

-0.67

 Bounce

-0.66

 Dawkins

-0.63

OSE

-0.62

fir

-0.61

 breasts

-0.60

POSITIVE LOGITS

orial

1.09

terness

0.93

ror

0.91

amen

0.86

abul

0.85

ept

0.85

atown

0.85

icians

0.84

itial

0.84

heastern

0.84

Activations Density 0.028%

words related to specific names or terms such as "Middlesbrough," "numeracy," and "Northeastern."

references to specific places and names, particularly in a statistical or comparative context

No Comments

No Known Activations