INDEX

Explanations

references to K-pop and its related terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

oser

-0.16

-0.15

Å¡e

-0.15

anka

-0.14

 Ð¾Ð±Ð»Ð°

-0.14

FRING

-0.14

 Govern

-0.14

åľ¨çº¿éĺħè¯»

-0.14

 Foley

-0.14

ÏĦÎ±Î¹

-0.14

POSITIVE LOGITS

 idol

0.26

 agency

0.23

 idols

0.23

 nuest

0.22

 agencies

0.20

TXT

0.19

aes

0.19

boy

0.19

 member

0.19

 Idol

0.19

Activations Density 0.043%

references to K-pop and its related terminology

No Comments

No Known Activations