INDEX

Explanations

references to theft and stolen items

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

mlin

-0.17

ROTO

-0.17

Segue

-0.15

ghi

-0.15

IMS

-0.14

utomation

-0.14

zano

-0.14

sponsor

-0.13

unker

-0.13

hydr

-0.13

POSITIVE LOGITS

 steal

0.61

 theft

0.60

 stealing

0.57

 stolen

0.57

 stole

0.56

 steals

0.56

pil

0.52

 Theft

0.50

 thief

0.46

 thieves

0.46

Activations Density 0.364%

references to theft and stolen items

No Comments

No Known Activations