INDEX

Explanations

keywords related to lids, covers, and closures

oai_token-act-pair · gpt-3.5-turbo

references to containers or covers, specifically lids

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

Ĥª

-0.75

issance

-0.74

merce

-0.72

fortune

-0.71

agents

-0.68

 Antar

-0.67

schild

-0.67

bidden

-0.67

 Borough

-0.65

 Bread

-0.63

POSITIVE LOGITS

lid

1.05

stones

0.98

ysis

0.94

stone

0.89

bell

0.88

plates

0.81

brow

0.80

wheel

0.77

 burner

0.76

wcs

0.75

Activations Density 0.022%

keywords related to lids, covers, and closures

references to containers or covers, specifically lids

No Comments

No Known Activations