INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 7-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L7R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.7.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

agas

-0.18

lobs

-0.15

legg

-0.15

experimental

-0.15

enas

-0.14

eldorf

-0.14

abra

-0.14

atoi

-0.14

chedulers

-0.14

aft

-0.14

POSITIVE LOGITS

_relations

0.15

zug

0.15

erable

0.14

tha

0.14

-Smith

0.14

DRV

0.14

Vin

0.14

vin

0.14

 coch

0.14

ÏĦÎµÏģÎ±

0.13

Activations Density 0.000%

No Known Activations

This feature has no known activations.