INDEX

Explanations

references to memories or honoring someone's memory

oai_token-act-pair · gpt-3.5-turbo

phrases and concepts related to memory

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

SHIP

-0.74

kef

-0.69

ats

-0.68

leys

-0.64

 Bundy

-0.64

Ã»

-0.64

ANY

-0.63

 Alaska

-0.59

obar

-0.59

ouk

-0.59

POSITIVE LOGITS

 memory

1.18

memory

1.03

 Memory

1.02

 remembrance

0.91

 Memories

0.90

 recollection

0.88

 memories

0.86

 retrieval

0.81

Memory

0.78

remember

0.75

Activations Density 0.014%

references to memories or honoring someone's memory

phrases and concepts related to memory

No Comments

No Known Activations