INDEX

Explanations

programming-related keywords and method definitions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

assen

-0.16

alles

-0.15

ninger

-0.15

arcy

-0.14

urette

-0.14

ector

-0.14

icrous

-0.14

iversit

-0.14

asar

-0.14

Ã©ment

-0.14

POSITIVE LOGITS

é¢Ħè§Ī

0.18

.qq

0.15

uko

0.14

 Hale

0.14

EEE

0.13

 nackte

0.13

oop

0.13

hta

0.13

 caps

0.13

zano

0.13

Activations Density 0.002%

programming-related keywords and method definitions

No Comments

No Known Activations