INDEX

Explanations

references to links and further information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

 Yates

-0.14

ÏīÎ½

-0.13

 Cousins

-0.13

 Unidos

-0.13

ham

-0.13

 Ø³Ùħ

-0.13

ãĤªãĥª

-0.13

ãĤ¦ãĤ¹

-0.13

Ĭ¶

-0.13

Ð¸ÑĤÐ°

-0.13

POSITIVE LOGITS

 jump

0.28

continue

0.25

Continue

0.25

 Continue

0.24

 continue

0.24

 continuation

0.21

 Jump

0.21

 Ø§Ø¯Ø§ÙħÙĩ

0.20

 break

0.20

 continues

0.20

Activations Density 0.032%

references to links and further information

No Comments

No Known Activations