INDEX

Explanations

vocabulary-related terms

oai_token-act-pair · gpt-3.5-turbo

terms related to vocational training and education

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

 Dragonbound

-0.75

 reconciliation

-0.68

 spotting

-0.67

 Territory

-0.65

 Siege

-0.65

 Sahara

-0.64

 Wonderland

-0.64

 Cunningham

-0.63

 guiActiveUnfocused

-0.63

 Dynasty

-0.63

POSITIVE LOGITS

abulary

1.18

iferation

1.03

ifer

1.00

aton

0.99

alled

0.97

ourses

0.95

ational

0.94

ials

0.93

learn

0.91

atism

0.90

Activations Density 0.031%

vocabulary-related terms

terms related to vocational training and education

No Comments

No Known Activations