INDEX

Explanations

references to historical events involving notable disasters and criminal activities

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

_INITIALIZER

-0.15

 Refugee

-0.15

 hazards

-0.15

rouw

-0.14

swick

-0.14

æģ

-0.14

 Parenthood

-0.14

jav

-0.14

 voksen

-0.14

ombat

-0.13

POSITIVE LOGITS

 serial

0.29

 Serial

0.28

 SERIAL

0.26

Serial

0.25

 profiler

0.24

 Zodiac

0.24

serial

0.24

 cold

0.23

 Bundy

0.22

Uns

0.22

Activations Density 0.017%

references to historical events involving notable disasters and criminal activities

No Comments

No Known Activations