INDEX

Explanations

proper nouns of specific locations or entities followed by the word "Place"

oai_token-act-pair · gpt-3.5-turbo

mentions of specific locations or places

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

reb

-0.71

hard

-0.70

 blown

-0.67

rieg

-0.66

 outl

-0.65

 branching

-0.65

atical

-0.65

adm

-0.65

mails

-0.64

 budgets

-0.63

POSITIVE LOGITS

 Place

4.01

Place

2.75

place

2.02

 place

1.94

 Places

1.70

 Avenue

1.18

places

1.17

placed

1.10

 Circle

1.08

 Plaza

1.06

Activations Density 0.022%

proper nouns of specific locations or entities followed by the word "Place"

mentions of specific locations or places

No Comments

No Known Activations