INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 17-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L17R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.17.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

 principal

-0.16

 thanks

-0.15

=explode

-0.15

 count

-0.15

principal

-0.14

gang

-0.14

 prof

-0.14

 gang

-0.14

POSITIVE LOGITS

umm

0.17

åĢį

0.15

iap

0.14

ìĪľ

0.14

embedded

0.14

 Mezi

0.14

eced

0.14

.trip

0.14

çĺ

0.13

eeper

0.13

Activations Density 0.000%

No Known Activations

This feature has no known activations.