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

ptal

-0.16

yll

-0.14

 Jordan

-0.14

.Cons

-0.14

 feather

-0.14

ÐµÐ»Ð¸

-0.14

 Feather

-0.14

kÃ³w

-0.14

cxx

-0.14

ython

-0.13

POSITIVE LOGITS

 Plugin

0.20

plugin

0.19

 plugin

0.19

Gear

0.18

 undef

0.17

DBObject

0.17

=cut

0.17

Plugin

0.17

-plugin

0.17

 Centre

0.16

Activations Density 0.000%

No Known Activations

This feature has no known activations.