INDEX

Explanations

mentions of the word "Burn" or variations of it

oai_token-act-pair · gpt-3.5-turbo

references to the term "Burn" and its variations in different contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

esan

-0.77

 hormone

-0.69

 invari

-0.69

om

-0.69

 paternal

-0.68

 orth

-0.67

GPS

-0.67

rait

-0.67

 Tsuk

-0.66

mer

-0.65

POSITIVE LOGITS

Burn

3.67

 Burning

2.92

 Burn

2.64

 burn

2.15

burning

1.93

 burns

1.92

burn

1.91

 Burns

1.87

 burning

1.80

 burner

1.79

Activations Density 0.027%

mentions of the word "Burn" or variations of it

references to the term "Burn" and its variations in different contexts

No Comments

No Known Activations