INDEX

Explanations

references to mathematical notation or concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

alem

-0.18

ackers

-0.16

ipple

-0.16

IPLE

-0.15

 multiple

-0.15

ugas

-0.14

aker

-0.14

tab

-0.14

 Party

-0.14

Party

-0.14

POSITIVE LOGITS

Ã¥r

0.16

gis

0.16

.gdx

0.16

roj

0.15

Î¯Î¸

0.15

ersiz

0.15

edla

0.15

.ERR

0.15

igne

0.14

 ÑĢÐ°ÑģÐ¿ÑĢÐ¾ÑģÑĤ

0.14

Activations Density 0.019%

references to mathematical notation or concepts

No Comments

No Known Activations