INDEX

Explanations

references to announcements or significant updates

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

vez

-0.16

ãĥ³ãĤ°

-0.15

aneous

-0.15

ÐµÑĤÑĮ

-0.15

ewe

-0.14

peare

-0.14

tea

-0.14

allet

-0.14

eydi

-0.14

den

-0.14

POSITIVE LOGITS

ihilation

0.31

ouncements

0.30

ivers

0.30

 Arbor

0.29

ouncing

0.28

ounced

0.27

ounce

0.26

ounces

0.25

iversary

0.24

ouncer

0.23

Activations Density 0.013%

references to announcements or significant updates

No Comments

No Known Activations