INDEX

Explanations

references to different empires

oai_token-act-pair · gpt-3.5-turbo

references to an "Empire" in various contexts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rint

-0.89

inventoryQuantity

-0.86

gettable

-0.80

matter

-0.78

electric

-0.74

WAYS

-0.73

esters

-0.73

quickShipAvailable

-0.72

yrics

-0.72

paces

-0.72

POSITIVE LOGITS

 Strikes

1.13

 Empire

0.95

 State

0.81

 outpost

0.74

 Builder

0.73

perial

0.73

wide

0.72

ocene

0.71

 Nero

0.70

 Awakens

0.70

Activations Density 0.011%

references to different empires

references to an "Empire" in various contexts

No Comments

No Known Activations