INDEX

Explanations

references to figures and notes within a technical or academic document

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 15-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L15R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.15.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

ertino

-0.15

atorium

-0.15

uary

-0.15

 Duffy

-0.15

yleft

-0.15

bab

-0.15

 Townsend

-0.14

bose

-0.14

ucc

-0.14

 Goth

-0.14

POSITIVE LOGITS

_PUS

0.15

-BEGIN

0.14

_unpack

0.14

zos

0.14

anst

0.14

ë²Į

0.14

@qq

0.14

à¸£à¸ģ

0.14

fos

0.14

 tá»ĩ

0.13

Activations Density 0.033%

references to figures and notes within a technical or academic document

No Comments

No Known Activations