INDEX

Explanations

phrases indicating the presence of various subjects or themes in a context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

cks

-0.17

aldi

-0.16

le

-0.16

lek

-0.15

ÂŃt

-0.15

 Jose

-0.14

ve

-0.14

zar

-0.14

 JosÃ©

-0.13

POSITIVE LOGITS

uch

0.18

Ã¡Å¾

0.17

legen

0.17

DMIN

0.17

cus

0.16

gen

0.16

asma

0.16

ItemSelected

0.16

ustin

0.16

-toggler

0.16

Activations Density 0.036%

phrases indicating the presence of various subjects or themes in a context

No Comments

No Known Activations