INDEX

Explanations

names of people or places in various contexts

oai_token-act-pair · gpt-3.5-turbo

specific names or identifiers related to people, places, or entities

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

IFrame

Link

•EggList

No Comments

Negative Logits

 Cassandra

-0.83

 Brist

-0.82

 clipboard

-0.71

 Carnage

-0.69

 Cass

-0.66

cel

-0.65

Cas

-0.65

 Hart

-0.64

 embr

-0.64

 Melissa

-0.64

POSITIVE LOGITS

1.55

Å«

1.37

uf

1.29

uv

1.27

uh

1.27

1.26

uz

1.24

uc

1.24

uer

1.23

ub

1.23

Activations Density 0.248%

names of people or places in various contexts

specific names or identifiers related to people, places, or entities

No Comments

No Known Activations