INDEX

Explanations

various types of content categories and filters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

oslav

-0.17

cko

-0.16

apt

-0.16

Ø±Ø³ÛĮ

-0.15

estone

-0.14

Ð»Ð°Ð²Ð°

-0.14

egas

-0.14

ÑģÑĸÐ¼

-0.14

ech

-0.13

innie

-0.13

POSITIVE LOGITS

 bist

0.15

og

0.15

 maxlen

0.15

caret

0.14

 Pref

0.14

makt

0.14

åį

0.14

ificate

0.14

PEAR

0.14

((↵

0.14

Activations Density 0.037%

various types of content categories and filters

No Comments

No Known Activations