INDEX

Explanations

references to films and cinematic projects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

aea

-0.17

ucid

-0.15

rai

-0.14

Ð»Ð¾Ð¼

-0.14

usher

-0.14

ece

-0.14

ãģĹãģĭ

-0.14

ãĤĩãģĨ

-0.14

hled

-0.14

vÄĽÅĻ

-0.13

POSITIVE LOGITS

 follows

0.41

 centers

0.37

 follow

0.35

 centres

0.33

 tell

0.33

 tells

0.31

 chron

0.31

 Follow

0.30

follow

0.30

 chronic

0.30

Activations Density 0.175%

references to films and cinematic projects

No Comments

No Known Activations