INDEX

Explanations

phrases indicating specific tasks or actions that are expected or intended to happen

oai_token-act-pair · gpt-3.5-turbo

phrases that express obligation or expectation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

quickShipAvailable

-0.77

fram

-0.68

Textures

-0.67

Cola

-0.66

 Nationwide

-0.64

 è£ıè¦ļéĨĴ

-0.64

ãĤ¦ãĤ¹

-0.64

DAQ

-0.63

raq

-0.62

Plus

-0.61

POSITIVE LOGITS

be

1.03

 arrive

0.92

 deliver

0.90

 kill

0.86

 finish

0.84

 assassinate

0.84

 accompany

0.83

 protect

0.83

 explode

0.83

acea

0.82

Activations Density 0.101%

phrases indicating specific tasks or actions that are expected or intended to happen

phrases that express obligation or expectation

No Comments

No Known Activations