INDEX

Explanations

names, specifically the name "Lisa"

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Lisa"

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

apeake

-0.73

hire

-0.72

ervative

-0.72

SHIP

-0.72

national

-0.71

grounds

-0.70

Ðº

-0.70

inventoryQuantity

-0.69

CLASSIFIED

-0.69

éŃĶ

-0.67

POSITIVE LOGITS

 Lamp

0.92

 Simpson

0.88

Sue

0.82

Rin

0.80

 Lisa

0.78

Joy

0.78

 Vander

0.77

anie

0.76

 Bloom

0.72

Tut

0.71

Activations Density 0.011%

names, specifically the name "Lisa"

mentions of the name "Lisa"

No Comments

No Known Activations