INDEX

Explanations

mentions of memorabilia or memorization-related concepts

oai_token-act-pair · gpt-3.5-turbo

references to memory and memorization

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

OHN

-0.72

IRO

-0.69

 Dani

-0.67

roads

-0.65

PO

-0.60

 Ames

-0.59

redd

-0.58

enthal

-0.58

Nah

-0.56

McF

-0.55

POSITIVE LOGITS

abilia

1.96

ably

1.29

andum

1.27

ization

1.12

izations

1.11

ized

1.11

isable

1.07

otype

1.07

ciating

1.07

izing

1.06

Activations Density 0.034%

mentions of memorabilia or memorization-related concepts

references to memory and memorization

No Comments

No Known Activations