INDEX

Explanations

named entities and titles related to media, culture, and awards

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

ceae

-0.15

 Mast

-0.15

 ÄĳÃ¢u

-0.15

ascus

-0.14

íı°

-0.14

Ĥæķ°

-0.14

ohan

-0.14

yscale

-0.14

tg

-0.13

 ÙħÙĦ

-0.13

POSITIVE LOGITS

âĢ¦the

0.14

ÛĮØ¯Ø§

0.14

-fw

0.13

'gc

0.13

iese

0.13

<*>

0.13

 nues

0.13

ÄįnÃŃk

0.13

0.12

zcze

0.12

Activations Density 0.215%

named entities and titles related to media, culture, and awards

No Comments

No Known Activations