INDEX

Explanations

numerical identifiers and bibliographic details related to publications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

oin

-0.14

fur

-0.14

eding

-0.14

affer

-0.13

enn

-0.13

ullan

-0.13

ansson

-0.13

Mgr

-0.13

ti

-0.13

rape

-0.13

POSITIVE LOGITS

NECT

0.17

 https

0.17

https

0.16

Aceptar

0.16

âĹİ

0.15

ereo

0.15

LOAT

0.15

eldo

0.15

INGTON

0.14

Activations Density 0.091%

numerical identifiers and bibliographic details related to publications

No Comments

No Known Activations