INDEX

Explanations

instances where actions are taken towards oneself

oai_token-act-pair · gpt-3.5-turbo

references to the concept of self-identity or self-reference

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ozy

-0.75

ories

-0.75

kefeller

-0.73

orian

-0.70

cemic

-0.70

actic

-0.69

yson

-0.68

oft

-0.67

ype

-0.67

opoly

-0.66

POSITIVE LOGITS

 herself

1.11

 proport

0.94

athed

0.91

 tremend

0.88

 conduc

0.84

 redes

0.83

selves

0.82

 perspect

0.81

self

0.80

 behavi

0.79

Activations Density 0.004%

instances where actions are taken towards oneself

references to the concept of self-identity or self-reference

No Comments

No Known Activations