INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

buat

-0.16

edBy

-0.16

beb

-0.16

stm

-0.15

@student

-0.14

 ullam

-0.14

:return

-0.14

 otel

-0.14

ìĥĿ

-0.14

heimer

-0.14

POSITIVE LOGITS

uner

0.16

appa

0.15

enu

0.14

 therein

0.14

 perhaps

0.14

 Atlas

0.14

0.13

Cad

0.13

Activations Density 0.516%

No Comments

No Known Activations