INDEX

Explanations

words related to locations, such as cities and states

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names of people and places

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

hex

-0.58

sect

-0.58

acity

-0.57

availability

-0.57

jet

-0.56

steel

-0.55

 reckoning

-0.55

 Berk

-0.55

 pretext

-0.54

 slides

-0.54

POSITIVE LOGITS

OUS

1.33

 EDITION

1.29

ING

1.24

URA

1.23

ED

1.23

ION

1.23

URES

1.21

SHIP

1.20

IAN

1.19

ILL

1.19

Activations Density 0.196%

words related to locations, such as cities and states

proper nouns, particularly names of people and places

No Comments

No Known Activations