INDEX

Explanations

terms related to processing and organization

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

iez

-0.20

Ã¼zel

-0.16

zenÃŃ

-0.16

Î±Î½Î¬

-0.16

fo

-0.15

ocht

-0.15

orz

-0.15

sku

-0.15

anean

-0.15

zyst

-0.15

POSITIVE LOGITS

ise

0.57

ises

0.50

ISE

0.45

ising

0.44

iser

0.41

ised

0.41

led

0.40

ISING

0.39

isers

0.36

ISED

0.33

Activations Density 0.104%

terms related to processing and organization

No Comments

No Known Activations