INDEX

Explanations

phrases related to user applications and goals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

jac

-0.17

Ã½t

-0.16

isz

-0.15

hee

-0.15

éĺ

-0.14

Observable

-0.14

fell

-0.14

.mapping

-0.14

\Unit

-0.14

adi

-0.13

POSITIVE LOGITS

 xong

0.23

emiz

0.20

 Porno

0.17

isphere

0.17

.mozilla

0.16

 Sesso

0.15

 PROCUREMENT

0.15

 ActionTypes

0.15

 prelim

0.15

Ä±ndÄ±r

0.14

Activations Density 0.158%

phrases related to user applications and goals

No Comments

No Known Activations