INDEX

Explanations

items of detailed identification related to official documents and codes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

Ø¤

-0.15

upy

-0.14

lia

-0.14

INGTON

-0.14

ugas

-0.14

ington

-0.14

ilo

-0.14

venta

-0.14

BUF

-0.13

heels

-0.13

POSITIVE LOGITS

ASI

0.17

Ð·Ð¼

0.15

aison

0.14

aisy

0.14

_SIMPLE

0.14

ause

0.14

anger

0.14

pole

0.13

.quant

0.13

 oldValue

0.13

Activations Density 0.033%

items of detailed identification related to official documents and codes

No Comments

No Known Activations