INDEX

Explanations

phrases related to dangers and ironies

oai_token-act-pair · gpt-3.5-turbo

phrases that discuss the concept of "dangers" or negative aspects of various topics

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

-0.76

CIA

-0.70

phabet

-0.69

uve

-0.68

ACP

-0.66

 istg

-0.64

 ILCS

-0.64

Ã±o

-0.64

Pol

-0.64

lem

-0.63

POSITIVE LOGITS

 these

0.78

 interconnected

0.70

 humankind

0.69

 hindsight

0.69

 sorts

0.68

 this

0.68

 mankind

0.68

 confronting

0.67

 nature

0.66

 balancing

0.66

Activations Density 0.183%

phrases related to dangers and ironies

phrases that discuss the concept of "dangers" or negative aspects of various topics

No Comments

No Known Activations