INDEX

Explanations

words related to programming or coding, specifically focusing on templates

oai_token-act-pair · gpt-3.5-turbo

terms related to templates in programming or documentation

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

ecause

-0.87

iqueness

-0.87

bringing

-0.85

better

-0.84

icult

-0.82

ells

-0.82

azar

-0.81

obbies

-0.81

ymes

-0.79

agher

-0.78

POSITIVE LOGITS

plates

1.05

 template

1.03

 Template

0.91

 templates

0.87

 markup

0.84

Url

0.83

 tags

0.76

tag

0.76

 html

0.73

 outline

0.73

Activations Density 0.014%

words related to programming or coding, specifically focusing on templates

terms related to templates in programming or documentation

No Comments

No Known Activations