INDEX

Explanations

references to popular video games and their characters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

 Kling

-0.19

 bande

-0.15

opot

-0.14

 Decompiled

-0.13

EK

-0.13

 Oklahoma

-0.13

.Selenium

-0.13

frey

-0.13

obil

-0.13

POSITIVE LOGITS

 Resident

0.20

 Disp

0.19

Disp

0.16

mons

0.16

ULE

0.16

 cabinet

0.16

lant

0.15

 chiá»ģu

0.15

uler

0.15

 resident

0.15

Activations Density 0.016%

references to popular video games and their characters

No Comments

No Known Activations