INDEX

Explanations

function-related keywords and structural components within programming code

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

aja

-0.15

aq

-0.15

 ControllerBase

-0.15

AQ

-0.14

 Cone

-0.14

 Ð±Ð¾Ñı

-0.14

Ø¨Ø¯

-0.14

ëī´ìĬ¤

-0.13

à¤ľà¤°

-0.13

.office

-0.13

POSITIVE LOGITS

ãĥ¼ãĥŀ

0.15

ande

0.15

orem

0.14

unken

0.14

apon

0.14

icut

0.14

iglia

0.14

gg

0.14

edi

0.13

dehyde

0.13

Activations Density 0.107%

function-related keywords and structural components within programming code

No Comments

No Known Activations