INDEX

Explanations

references to art exhibitions and artist techniques

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Designs

-0.17

arto

-0.17

rint

-0.15

 Fitzgerald

-0.15

 mosaic

-0.15

rof

-0.14

 fundra

-0.14

urtles

-0.14

specialchars

-0.14

haf

-0.13

POSITIVE LOGITS

vit

0.20

 installation

0.18

 gallery

0.17

 hung

0.17

viewer

0.16

Untitled

0.16

 prepar

0.16

gallery

0.16

łí

0.16

lep

0.15

Activations Density 0.045%

references to art exhibitions and artist techniques

No Comments

No Known Activations