INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

 otherwise

-0.19

 later

-0.19

 neither

-0.17

otherwise

-0.16

 earlier

-0.16

later

-0.16

 dnes

-0.16

 Ð·Ð°Ð²ÑĤÑĢÐ°

-0.14

 originally

-0.14

ìĹĨìĿĮ

-0.14

POSITIVE LOGITS

 ÙħÛĮÙĦØ§Ø¯ÛĮ

0.20

ãĤĪãģĨ

0.16

 Ð³Ð¾Ð´Ñĥ

0.16

 ÑĢÐ¾ÑĨÑĸ

0.16

 erfol

0.15

rine

0.15

-present

0.15

 werd

0.14

ÙħÛĮÙĦ

0.14

.au

0.14

Activations Density 0.067%

No Comments

No Known Activations