INDEX

Explanations

commands and structure related to programming code

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

 detr

-0.16

 Kush

-0.13

_READONLY

-0.12

 ÃĤ

-0.12

à¥ĭà¤ķà¤°

-0.12

[url

-0.12

ä¸Ī

-0.12

amac

-0.12

 Buddh

-0.12

 darken

-0.12

POSITIVE LOGITS

MethodImpl

0.14

Mathf

0.14

&&&&

0.14

adle

0.13

ogenesis

0.13

iola

0.13

 Gross

0.13

ÅĤa

0.13

oriously

0.13

ynet

0.13

Activations Density 0.010%

commands and structure related to programming code

No Comments

No Known Activations