INDEX

Explanations

mentions of a specific location named "Cal" followed by a string of text

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 favour

-0.64

 depreciation

-0.62

lihood

-0.62

 Myster

-0.61

 Lans

-0.60

 accord

-0.60

 Ajax

-0.59

beh

-0.58

 praise

-0.58

 chosen

-0.58

POSITIVE LOGITS

houn

1.28

endars

1.21

culus

1.19

cium

1.15

iphate

1.10

gary

1.09

culation

1.08

iber

1.07

guns

1.02

abre

1.01

Activations Density 0.026%

mentions of a specific location named "Cal" followed by a string of text

No Comments

No Known Activations