INDEX

Explanations

references to specific time periods in history

oai_token-act-pair · gpt-3.5-turbo

references to the passage of time, specifically focusing on the term "years ago."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ozyg

-0.64

iffe

-0.63

aunch

-0.63

agonist

-0.62

inburgh

-0.62

ateral

-0.61

malink

-0.60

ocker

-0.60

own

-0.60

 collateral

-0.60

POSITIVE LOGITS

ago

1.07

osaurs

0.76

Ago

0.75

 hindsight

0.70

 olds

0.70

idia

0.68

BCE

0.68

glass

0.67

âĢİ

0.67

fu

0.66

Activations Density 0.061%

references to specific time periods in history

references to the passage of time, specifically focusing on the term "years ago."

No Comments

No Known Activations