INDEX

Explanations

instances of the word "last" followed by a number

oai_token-act-pair · gpt-3.5-turbo

instances of the word "last" and its context within statements referencing time

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

holes

-0.73

keye

-0.67

adle

-0.66

oute

-0.65

andise

-0.63

okers

-0.63

ordes

-0.63

arrass

-0.63

aunders

-0.62

asure

-0.61

POSITIVE LOGITS

 week

1.31

 month

1.30

 year

1.19

 night

1.06

 summer

1.00

 semester

0.97

 weekend

0.94

decade

0.94

 Month

0.94

 autumn

0.92

Activations Density 0.063%

instances of the word "last" followed by a number

instances of the word "last" and its context within statements referencing time

No Comments

No Known Activations