INDEX

Explanations

acronyms, abbreviations, or initialisms related to organizations or programs

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

 VerfÃ¼g

-0.16

DD

-0.16

ythe

-0.15

cht

-0.14

 scaleY

-0.14

chest

-0.14

nost

-0.14

Ø§ØªÙĩ

-0.13

 applicationContext

-0.13

PLUS

-0.13

POSITIVE LOGITS

dojo

0.16

å¡ļ

0.16

eten

0.15

èĳĹ

0.14

oscopic

0.14

skou

0.14

yii

0.14

rych

0.14

Rae

0.14

dock

0.14

Activations Density 0.286%

acronyms, abbreviations, or initialisms related to organizations or programs

No Comments

No Known Activations