INDEX

Explanations

references to the Dalai Lama

oai_token-act-pair · gpt-3.5-turbo

references to influential leaders, specifically the Dalai Lama and Lenin

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

hire

-0.82

reek

-0.81

ties

-0.74

IED

-0.74

ptroller

-0.71

tes

-0.68

EAR

-0.67

aved

-0.67

 Rangers

-0.67

tle

-0.65

POSITIVE LOGITS

 Lama

1.09

 Jinping

0.95

EStream

0.84

utra

0.82

onite

0.78

 seiz

0.78

eus

0.77

istani

0.76

DragonMagazine

0.75

 Patriarch

0.74

Activations Density 0.013%

references to the Dalai Lama

references to influential leaders, specifically the Dalai Lama and Lenin

No Comments

No Known Activations