INDEX

Explanations

variables being assigned or initialized in a programming context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

ymm

-0.17

.gf

-0.15

utr

-0.15

 bump

-0.14

usta

-0.14

ucu

-0.14

ynet

-0.14

ampp

-0.14

permissions

-0.14

LOWER

-0.14

POSITIVE LOGITS

oke

0.16

lane

0.14

bung

0.14

ICO

0.14

rip

0.14

och

0.14

isser

0.14

wyn

0.14

unately

0.13

lei

0.13

Activations Density 0.118%

variables being assigned or initialized in a programming context

No Comments

No Known Activations