INDEX

Explanations

frequency-related terms and expressions emphasizing quantity and repetition

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L1R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.1.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

ulis

-0.15

 Pent

-0.15

enko

-0.15

ÏģÏĮ

-0.14

oz

-0.13

coni

-0.13

zc

-0.13

ala

-0.13

ãĤ¹ãĥ¬

-0.13

egment

-0.13

POSITIVE LOGITS

vak

0.16

plnÄĽ

0.14

arella

0.13

atically

0.13

bip

0.13

 Kann

0.13

Cro

0.13

tum

0.13

 mast

0.13

 continents

0.13

Activations Density 0.578%

frequency-related terms and expressions emphasizing quantity and repetition

No Comments

No Known Activations