INDEX

Explanations

intensifiers and modifiers that emphasize the degree of certain adjectives or descriptions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 7-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L7R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.7.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

ombat

-0.16

elix

-0.15

åī£

-0.15

 ÙĪØ§ÙĦØ³

-0.15

 Ð²ÐµÐ´ÑĮ

-0.14

Ð¾Ð±ÑĢÐ°Ð·

-0.14

.Dom

-0.14

eyen

-0.14

Î²Î¿

-0.14

agues

-0.13

POSITIVE LOGITS

asso

0.18

ored

0.14

imple

0.14

 ná»Ļi

0.14

 dere

0.14

bury

0.14

aws

0.14

arian

0.13

æĿĲ

0.13

icom

0.13

Activations Density 0.090%

intensifiers and modifiers that emphasize the degree of certain adjectives or descriptions

No Comments

No Known Activations