INDEX

Explanations

references to individuals or possessive pronouns

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.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

Ð¾Ð±Ð¸

-0.17

á»Ļi

-0.15

shal

-0.15

 SKIP

-0.15

iband

-0.15

voy

-0.14

pike

-0.14

Ð½Ð¾ÑģÑı

-0.14

ipel

-0.14

ãĤ¡

-0.14

POSITIVE LOGITS

èĬĿ

0.15

minated

0.15

Ã«

0.15

 station

0.14

azer

0.14

IST

0.14

chia

0.14

estre

0.13

azÄĥ

0.13

 é¹

0.13

Activations Density 0.007%

references to individuals or possessive pronouns

No Comments

No Known Activations