INDEX

Explanations

instances of the word "memories"

oai_token-act-pair · gpt-3.5-turbo

references to memories and reminiscence

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

SHIP

-0.73

yright

-0.68

atism

-0.67

Nich

-0.66

annis

-0.65

otomy

-0.65

OPLE

-0.63

osi

-0.62

retty

-0.61

 Scandinavian

-0.61

POSITIVE LOGITS

 Memories

1.44

 memories

1.33

 recollection

0.97

 memory

0.94

 Memory

0.93

 remembrance

0.89

 remembered

0.86

memory

0.85

remember

0.84

 Remem

0.83

Activations Density 0.010%

instances of the word "memories"

references to memories and reminiscence

No Comments

No Known Activations