INDEX

Explanations

references to film production and film-related terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

rend

-0.17

dru

-0.15

ounge

-0.15

esch

-0.14

retweeted

-0.14

Rat

-0.14

 tertiary

-0.14

_tran

-0.14

adio

-0.14

 Fever

-0.13

POSITIVE LOGITS

 silent

0.35

 Silent

0.30

silent

0.29

 Edison

0.27

 sessiz

0.25

 motion

0.24

 Thanh

0.23

 Motion

0.23

motion

0.23

Lum

0.22

Activations Density 0.064%

references to film production and film-related terminology

No Comments

No Known Activations