INDEX

Explanations

phrases related to causality or reason

oai_token-act-pair · gpt-3.5-turbo

phrases that introduce explanations or justifications

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

Contact

-0.88

LAB

-0.83

marine

-0.77

zona

-0.75

contact

-0.71

Movie

-0.70

Minimum

-0.68

BuyableInstoreAndOnline

-0.67

Dro

-0.66

Jr

-0.64

POSITIVE LOGITS

 instance

1.24

gotten

1.22

bidden

1.19

 example

1.15

 centuries

1.09

 millennia

0.97

cing

0.95

 reasons

0.91

 decades

0.91

cible

0.90

Activations Density 0.096%

phrases related to causality or reason

phrases that introduce explanations or justifications

No Comments

No Known Activations