INDEX

Explanations

punctuation and numerical symbols, especially those indicating relationships or proportions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.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

asma

-0.17

ottle

-0.15

NotAllowed

-0.14

hong

-0.14

aval

-0.14

edd

-0.14

 ÙħÛĮÙĦ

-0.13

amine

-0.13

nam

-0.13

Pager

-0.13

POSITIVE LOGITS

 ÐĴÐ¸ÐºÐ¸

0.16

Ð¶Ñĥ

0.16

cken

0.14

WON

0.14

 Ellison

0.13

Won

0.13

 ì°

0.13

liers

0.13

azon

0.13

 suits

0.13

Activations Density 0.260%

punctuation and numerical symbols, especially those indicating relationships or proportions

No Comments

No Known Activations