INDEX

Explanations

references to technical functions and functionalities within software or system processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 23-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.23.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

tar

-0.16

 Prest

-0.16

 Bender

-0.14

 pron

-0.14

ecc

-0.14

 uptake

-0.14

ARTH

-0.14

yna

-0.14

 Tubes

-0.14

uito

-0.13

POSITIVE LOGITS

.dw

0.27

 WINAPI

0.24

apis

0.23

Win

0.21

.cb

0.20

 SetLastError

0.20

_NT

0.19

.Quad

0.19

.lp

0.19

Win

0.19

Activations Density 0.057%

references to technical functions and functionalities within software or system processes

No Comments

No Known Activations