INDEX

Explanations

mentions of museums or specific museum names

oai_token-act-pair · gpt-3.5-turbo

references to various museums

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

cially

-0.78

 blockers

-0.69

 hashing

-0.67

 chains

-0.65

Pos

-0.64

lying

-0.63

 shif

-0.63

 crunch

-0.63

 clauses

-0.61

ass

-0.61

POSITIVE LOGITS

 Museum

3.86

 museum

2.44

 museums

2.16

useum

2.03

 Smithsonian

1.82

 Muse

1.67

 Library

1.47

 Archives

1.41

sonian

1.40

 Cemetery

1.35

Activations Density 0.013%

mentions of museums or specific museum names

references to various museums

No Comments

No Known Activations