INDEX

Explanations

references to viral content or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

 Podcast

-0.16

Pod

-0.15

bilt

-0.15

ocal

-0.14

odcast

-0.14

erli

-0.14

 defe

-0.14

 rolled

-0.14

 roll

-0.14

roll

-0.14

POSITIVE LOGITS

vir

0.46

 viral

0.43

Vir

0.40

VIR

0.36

vir

0.34

Vir

0.32

 circulated

0.30

 circulation

0.28

 circulating

0.28

Mem

0.24

Activations Density 0.041%

references to viral content or events

No Comments

No Known Activations