INDEX

Explanations

numeric values and their contextual relationships in a temporal or sequential format

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

 Ð´Ð¾ÐºÑĥÐ¼

-0.23

 ÑģÐ¾Ð±Ñĭ

-0.22

 Ð¸Ð½Ð´Ð¸Ð²Ð¸Ð´Ñĥ

-0.21

 Ð¸Ð½ÑĦÐ¾ÑĢÐ¼Ð°

-0.20

 ÑģÑĥÑīÐµ

-0.17

 ÐµÑģÑĤÐµ

-0.17

 ÑģÐ¾ÑħÑĢÐ°

-0.16

å¤ļãģĦ

-0.16

aklÄ±

-0.16

 ÑĤÐ°Ð±Ð»Ð¸

-0.15

POSITIVE LOGITS

 Ð·Ð°ÑģÐ¾Ð±

0.19

 ÑģÑĤÐ°Ð»

0.16

ÑĨÑĭ

0.16

dio

0.16

 ÑģÑĦ

0.16

 ÑģÐ¼Ð¾Ð³

0.15

 Ð·Ð°ÑģÑĤ

0.15

chal

0.15

 wurde

0.15

Ã¢u

0.15

Activations Density 0.024%

numeric values and their contextual relationships in a temporal or sequential format

No Comments

No Known Activations