INDEX

Explanations

proper nouns and specific terms related to technology or products

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

Els

-1.02

heng

-0.85

 Henri

-0.83

oken

-0.79

MON

-0.79

aldehyde

-0.78

EL

-0.77

egal

-0.75

anchester

-0.74

llah

-0.74

POSITIVE LOGITS

Sp

2.48

Sp

2.36

sp

2.15

sp

2.03

SP

1.86

SP

1.66

 Spark

1.55

 Spur

1.51

 Spin

1.45

Spy

1.43

Activations Density 0.255%

proper nouns and specific terms related to technology or products

No Comments

No Known Activations