INDEX

Explanations

phrases related to specific projects or organizations

oai_token-act-pair · gpt-3.5-turbo

repeated phrases or sequences that emphasize importance or significance

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

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

 oven

-0.85

 nuts

-0.77

 precaution

-0.74

umin

-0.72

 proced

-0.70

 protective

-0.69

orts

-0.68

 palm

-0.67

 hemor

-0.67

 Lumpur

-0.66

POSITIVE LOGITS

advertisement

1.02

along

0.99

meaning

0.98

feat

0.96

something

0.95

among

0.95

––

0.95

perhaps

0.94

âĢ¢âĢ¢

0.94

particularly

0.93

Activations Density 0.047%

phrases related to specific projects or organizations

repeated phrases or sequences that emphasize importance or significance

No Comments

No Known Activations