INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 28-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L28R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.28.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

loquent

-0.16

ucid

-0.15

LOCKS

-0.15

 Epid

-0.15

bon

-0.15

ufac

-0.14

Tank

-0.14

enheim

-0.14

ensch

-0.14

SingleOrDefault

-0.14

POSITIVE LOGITS

 rather

0.23

 tend

0.19

rather

0.19

 tends

0.18

 Rather

0.17

 tendency

0.17

 mostly

0.16

 ways

0.16

 manner

0.16

 Mostly

0.16

Activations Density 0.000%

No Known Activations

This feature has no known activations.