INDEX

Explanations

affirmative statements and indicators of truthfulness

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

Embed

-0.15

bak

-0.15

dle

-0.15

ÎŃÏģÎ³

-0.14

alle

-0.14

DataTask

-0.14

.scalablytyped

-0.14

opsy

-0.13

ÐµÐ¿ÑĤÐ¸

-0.13

POSITIVE LOGITS

 equally

0.21

 also

0.17

also

0.17

ä¹Łæľī

0.16

UGIN

0.15

ynet

0.15

à¸§à¸Ķ

0.15

 ALSO

0.15

ä¹Ł

0.15

ê¸°ëıĦ

0.15

Activations Density 0.071%

affirmative statements and indicators of truthfulness

No Comments

No Known Activations