INDEX

Explanations

punctuation marks and symbols

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

 McCorm

-0.17

ett

-0.16

ipar

-0.15

ies

-0.15

 Mall

-0.15

rouw

-0.14

PIL

-0.14

esel

-0.14

LECT

-0.14

 loop

-0.14

POSITIVE LOGITS

public

0.23

 class

0.22

class

0.22

 public

0.19

ami

0.16

atore

0.16

:class

0.15

 sÄ±nÄ±f

0.15

aclass

0.15

	class

0.15

Activations Density 0.012%

punctuation marks and symbols

No Comments

No Known Activations