INDEX

Explanations

No Explanations Found

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

Ø³ÙĪ

-0.15

 Bulk

-0.15

enko

-0.15

lay

-0.14

UD

-0.14

oya

-0.14

 kolo

-0.14

Ã¶ff

-0.14

Bulk

-0.14

ocaust

-0.14

POSITIVE LOGITS

.twig

0.16

ant

0.16

.obtain

0.15

 expl

0.15

áº©u

0.15

opping

0.15

 apart

0.15

rec

0.15

urer

0.14

taj

0.14

Activations Density 0.000%

No Known Activations

This feature has no known activations.