INDEX

Explanations

names and titles related to professions and occupations

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

vise

-0.16

lagen

-0.15

á»Ń

-0.15

 unnamed

-0.14

atak

-0.14

ÑĥÑģÑĤÐ°

-0.14

.Produ

-0.14

posables

-0.14

ãģ»ãģ©

-0.13

eks

-0.13

POSITIVE LOGITS

 active

0.22

who

0.20

active

0.17

 aktiv

0.17

Inactive

0.16

 mostly

0.16

who

0.16

 Born

0.15

 Active

0.15

 ÑĢÐ¾Ð´

0.15

Activations Density 0.034%

names and titles related to professions and occupations

No Comments

No Known Activations