INDEX

Explanations

terms related to outlaws and fighters, particularly focusing on ninja characters

oai_token-act-pair · gpt-3.5-turbo

terms related to ninjas and outlaws

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

ween

-0.89

bow

-0.82

eries

-0.80

agall

-0.78

ibrary

-0.78

eus

-0.78

orius

-0.77

inus

-0.76

eele

-0.76

ochond

-0.76

POSITIVE LOGITS

 warrior

0.82

 warriors

0.81

 Cowboy

0.73

 mercenary

0.73

 Warrior

0.69

 boxing

0.68

Sold

0.67

 pirate

0.66

 Commando

0.66

Fighting

0.65

Activations Density 0.040%

terms related to outlaws and fighters, particularly focusing on ninja characters

terms related to ninjas and outlaws

No Comments

No Known Activations