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

-0.14

 behavioural

-0.14

bounce

-0.14

lyph

-0.14

leyen

-0.13

uffle

-0.13

PRINTF

-0.13

 programme

-0.13

nodoc

-0.13

POSITIVE LOGITS

 fucking

0.26

fuck

0.25

 fuck

0.25

 fucked

0.23

 Fucking

0.23

 FUCK

0.20

 fucks

0.20

Fuck

0.20

 Fuck

0.20

 cunt

0.19

Activations Density 0.000%

No Known Activations

This feature has no known activations.