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

mÄĽ

-0.17

 addCriterion

-0.16

iores

-0.15

osit

-0.15

MaxY

-0.14

ustos

-0.14

Dow

-0.14

lick

-0.14

eneg

-0.14

chÃ©

-0.14

POSITIVE LOGITS

Ru

0.22

-up

0.20

-U

0.19

upy

0.18

up

0.18

UP

0.17

ucker

0.17

_up

0.17

Vu

0.17

up

0.16

Activations Density 0.000%

No Known Activations

This feature has no known activations.