INDEX

Explanations

phrases indicating possession or affiliation with someone or something

oai_token-act-pair · gpt-3.5-turbo

references to personal ownership or self-identity

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

hesda

-0.79

antle

-0.74

xual

-0.74

romeda

-0.74

anwhile

-0.73

olic

-0.72

MER

-0.71

obar

-0.71

oms

-0.70

 ILCS

-0.69

POSITIVE LOGITS

 personal

0.84

 accord

0.83

 backyard

0.78

 version

0.76

 affairs

0.75

 hars

0.74

 selves

0.74

 creations

0.73

 admission

0.72

 downfall

0.70

Activations Density 0.037%

phrases indicating possession or affiliation with someone or something

references to personal ownership or self-identity

No Comments

No Known Activations