INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

jvu

-0.18

bett

-0.16

FirstResponder

-0.15

ATAR

-0.15

ventus

-0.14

acket

-0.14

Ì£

-0.14

atar

-0.14

_capabilities

-0.14

istogram

-0.14

POSITIVE LOGITS

ioni

0.15

apy

0.15

Swe

0.14

 Guild

0.13

 |↵

0.13

 ÎĵÎ¿Ïħ

0.13

harma

0.13

Ess

0.13

elta

0.13

Activations Density 0.000%

No Known Activations

This feature has no known activations.