INDEX

Explanations

references to time periods within the year, specifically the second half

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

urat

-0.64

mercial

-0.64

ashtra

-0.63

incinn

-0.62

ilial

-0.59

ographed

-0.58

enegger

-0.58

SHIP

-0.57

tained

-0.57

ograp

-0.57

POSITIVE LOGITS

enium

0.95

pipe

0.92

dozen

0.84

century

0.80

imester

0.79

hour

0.77

wich

0.76

 dozen

0.75

teenth

0.75

finals

0.74

Activations Density 0.030%

references to time periods within the year, specifically the second half

No Comments

No Known Activations