INDEX

Explanations

conditional phrases or statements

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

Ð¾ÐºÐ¾Ð»

-0.18

uÅŁ

-0.17

ohana

-0.16

ounters

-0.15

:UIAlert

-0.14

ãģ¨ãģĨ

-0.14

erland

-0.14

-prepend

-0.13

WithValue

-0.13

ationToken

-0.13

POSITIVE LOGITS

 anything

0.32

 ever

0.30

 anyone

0.29

 anybody

0.27

any

0.26

 nothing

0.25

 only

0.25

 memory

0.25

tek

0.24

 imitation

0.23

Activations Density 0.073%

conditional phrases or statements

No Comments

No Known Activations