INDEX

Explanations

code-related syntax and function structures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

 Ukra

-0.15

shortcode

-0.15

ymi

-0.15

ç¿¼

-0.14

(æĹ¥

-0.14

 DÃ¼z

-0.14

braco

-0.14

æľĭ

-0.14

Ð»ÑıÐ´

-0.14

itched

-0.14

POSITIVE LOGITS

 done

0.22

done

0.20

(done

0.20

 Done

0.19

igu

0.19

 ^{↵

0.17

isol

0.17

()

0.17

.done

0.16

 DONE

0.16

Activations Density 0.008%

code-related syntax and function structures

No Comments

No Known Activations