INDEX

Explanations

mentions of military ranks, units, and assignments

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

parency

-0.78

 refere

-0.78

woo

-0.78

 clauses

-0.75

rils

-0.74

 shaming

-0.73

 bloggers

-0.73

 cures

-0.73

 tipping

-0.73

nect

-0.71

POSITIVE LOGITS

 Infantry

1.79

 Regiment

1.67

 Battalion

1.61

 Squadron

1.57

 infantry

1.56

 Commando

1.48

 battalion

1.43

 Army

1.42

 Marines

1.41

 Corps

1.40

Activations Density 0.347%

mentions of military ranks, units, and assignments

No Comments

No Known Activations