INDEX

Explanations

terms related to assistance or help in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

èŃ

-0.16

rita

-0.15

tha

-0.15

mand

-0.15

ungi

-0.15

 mand

-0.15

opr

-0.15

Ð»ÐµÑĩ

-0.14

ione

-0.14

Ø§

-0.14

POSITIVE LOGITS

 navigation

0.29

navigate

0.27

 navigate

0.27

 navigating

0.26

 navig

0.25

navigation

0.25

Navigate

0.24

 Navigation

0.23

-navigation

0.22

nav

0.21

Activations Density 0.183%

terms related to assistance or help in various contexts

No Comments

No Known Activations