INDEX

Explanations

occurrences of the word "class" as well as various linguistic and structural elements that signal classification or categorization within text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

ÑĩÐµ

-0.17

ÑĪÐºÑĥ

-0.16

ÐµÐºÑģÐ¸

-0.15

enskÃ©

-0.15

QRST

-0.14

ulus

-0.14

apa

-0.14

idebar

-0.13

Atl

-0.13

ius

-0.13

POSITIVE LOGITS

zem

0.18

lein

0.16

Ø¨ÙĪØ±

0.15

PullParser

0.15

apur

0.14

ĶåĽŀ

0.14

lando

0.14

affer

0.14

Ð¸ÐºÑĥ

0.14

 Ð´ÐµÐ»

0.14

Activations Density 0.002%

occurrences of the word "class" as well as various linguistic and structural elements that signal classification or categorization within text

No Comments

No Known Activations