INDEX

Explanations

parts of programming code and syntax elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 10-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L10R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.10.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

ÑĤÐµÑĢÐ½

-0.15

&R

-0.14

punk

-0.14

precated

-0.13

lander

-0.13

emet

-0.13

 Morse

-0.13

olf

-0.13

ÑĪÐºÐ°

-0.13

ATAB

-0.13

POSITIVE LOGITS

eck

0.18

afil

0.16

atts

0.15

ç´Ģ

0.15

umann

0.14

strup

0.14

æ¿

0.14

ÐµÐºÑĤÐ¸Ð²

0.14

osi

0.14

afort

0.14

Activations Density 0.073%

parts of programming code and syntax elements

No Comments

No Known Activations