INDEX

Explanations

proper nouns or names, especially those related to various locations, people or organizations

oai_token-act-pair · gpt-3.5-turbo

words related to geographical locations or proper nouns

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

arcity

-0.86

etheless

-0.75

atform

-0.75

 conflic

-0.71

 carbohyd

-0.69

ccording

-0.68

perature

-0.67

 unnecess

-0.65

ADRA

-0.65

 unden

-0.65

POSITIVE LOGITS

cloth

0.83

BSD

0.77

thal

0.76

 Hills

0.76

weed

0.75

wine

0.73

 Rite

0.73

velt

0.72

oft

0.72

SSL

0.69

Activations Density 0.381%

proper nouns or names, especially those related to various locations, people or organizations

words related to geographical locations or proper nouns

No Comments

No Known Activations