INDEX

Explanations

mentions of receiving or giving advice

oai_token-act-pair · gpt-3.5-turbo

references to giving or receiving advice

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

bang

-0.74

lock

-0.67

anne

-0.63

oor

-0.63

ching

-0.63

locks

-0.63

dimension

-0.62

FACE

-0.62

Jump

-0.62

ppa

-0.61

POSITIVE LOGITS

 Advice

1.08

 advice

1.05

 advising

1.02

 advised

0.93

 advise

0.91

 advises

0.78

 recommendation

0.77

 paralysis

0.76

uyomi

0.75

urai

0.74

Activations Density 0.030%

mentions of receiving or giving advice

references to giving or receiving advice

No Comments

No Known Activations