INDEX

Explanations

specific references to museums

oai_token-act-pair · gpt-3.5-turbo

mentions of museums

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

nir

-0.81

LAN

-0.67

shaw

-0.65

 Wast

-0.62

Leary

-0.61

iated

-0.61

maxwell

-0.61

lees

-0.60

 modem

-0.60

 speeds

-0.60

POSITIVE LOGITS

ĸļ

1.14

useum

1.01

 curator

0.93

arium

0.87

 museums

0.87

 museum

0.86

istry

0.83

 exhibits

0.81

 galleries

0.80

conservancy

0.76

Activations Density 0.020%

specific references to museums

mentions of museums

No Comments

No Known Activations