INDEX

Explanations

references to various practices or behaviors

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

mentions of the word "practice."

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

onge

-0.76

gin

-0.74

 panc

-0.72

aman

-0.72

aughter

-0.67

oor

-0.66

aline

-0.65

inki

-0.64

eele

-0.64

arger

-0.64

POSITIVE LOGITS

 practice

1.02

 Practices

0.94

 Practice

0.91

 practiced

0.88

 practices

0.86

 practise

0.86

 practicing

0.85

 pract

0.82

practice

0.81

 practitioners

0.77

Activations Density 0.022%

references to various practices or behaviors

mentions of the word "practice."

No Comments

No Known Activations