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

ileo

-0.15

duy

-0.15

 PMID

-0.15

stru

-0.14

ÑĤÑĢÐ¸

-0.14

VICES

-0.14

&T

-0.13

/browse

-0.13

prox

-0.13

POSITIVE LOGITS

otos

0.15

ETO

0.15

PlainText

0.15

 lots

0.15

rab

0.14

kor

0.14

èµĽ

0.14

ÐµÐ³Ð¾Ð´Ð½Ñı

0.14

EFR

0.14

antity

0.14

Activations Density 0.000%

No Known Activations

This feature has no known activations.