INDEX

Explanations

references or citations in the text

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

Ø§ÙĨÙĩ

-0.14

unce

-0.14

zik

-0.14

tement

-0.14

anse

-0.14

 ká»³

-0.14

asma

-0.14

tte

-0.13

 Cloth

-0.13

 caps

-0.13

POSITIVE LOGITS

 ÄĳoÃłn

0.16

ãĤ¹ãĤ³

0.14

NECTION

0.13

à¤¾à¤®à¤Ĺ

0.13

ateur

0.13

 Bench

0.13

 Franken

0.13

ateurs

0.13

andler

0.13

 ìłĦìļ©

0.13

Activations Density 0.002%

references or citations in the text

No Comments

No Known Activations