INDEX

Explanations

phrases indicating past experiences and transformations of places or identities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

 currently

-0.23

 soon

-0.22

çĽ®åīį

-0.21

soon

-0.19

 ìķĦì§ģ

-0.19

currently

-0.19

 presently

-0.18

 zatÃŃm

-0.18

inks

-0.17

yet

-0.17

POSITIVE LOGITS

(before

0.17

ãģĤãģ£ãģŁ

0.17

adays

0.16

 Ð²Ð²Ð°Ð¶

0.16

theless

0.16

à¹Ģà¸Ħà¸¢

0.16

aeda

0.16

/current

0.15

-before

0.15

 simply

0.15

Activations Density 0.114%

phrases indicating past experiences and transformations of places or identities

No Comments

No Known Activations