INDEX

Explanations

descriptions of architectural features and historical details

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 23-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.23.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

lient

-0.16

 hamburger

-0.15

Wyn

-0.15

amble

-0.15

 ctxt

-0.15

 Ð¿Ð»Ñİ

-0.14

Æ°Æ¡ng

-0.13

 Packing

-0.13

 ballistic

-0.13

èŃľ

-0.13

POSITIVE LOGITS

 Linden

0.18

 bend

0.16

 ensemble

0.15

Æ°á»£t

0.14

POSITE

0.14

 Administrative

0.14

 onUpdate

0.14

iso

0.14

 Relief

0.14

 spatial

0.13

Activations Density 0.115%

descriptions of architectural features and historical details

No Comments

No Known Activations