INDEX

Explanations

keywords related to military and warfare, specifically referring to "Allied" forces

oai_token-act-pair · gpt-3.5-turbo

references to the Allied forces and their operations during historical conflicts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

à

-0.89

BOOK

-0.88

inventoryQuantity

-0.81

Hop

-0.81

anity

-0.78

Minecraft

-0.78

aminer

-0.78

Book

-0.77

displayText

-0.74

hots

-0.74

POSITIVE LOGITS

 Allied

1.02

 Forces

0.95

 Allies

0.92

 parach

0.87

ãĤ¤ãĥĪ

0.81

 Powers

0.79

 Compos

0.76

Thy

0.76

 servic

0.75

 sympath

0.74

Activations Density 0.022%

keywords related to military and warfare, specifically referring to "Allied" forces

references to the Allied forces and their operations during historical conflicts

No Comments

No Known Activations