INDEX

Explanations

mentions of different oceans, particularly the Pacific Ocean, in the text

oai_token-act-pair · gpt-3.5-turbo

mentions of specific oceanic locations

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

<+

-0.71

ts

-0.69

 stiffness

-0.66

xual

-0.66

lik

-0.66

CRIP

-0.64

dos

-0.64

gh

-0.63

==

-0.63

 hasht

-0.62

POSITIVE LOGITS

 Ocean

3.96

Ocean

3.12

 ocean

2.11

 Seas

1.68

cean

1.64

 oceans

1.55

ceans

1.53

Sea

1.49

 Atlantic

1.45

 Neptune

1.38

Activations Density 0.012%

mentions of different oceans, particularly the Pacific Ocean, in the text

mentions of specific oceanic locations

No Comments

No Known Activations