INDEX

Explanations

colons followed by non-zero values

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

ovsky

-0.17

(æ°´

-0.15

andy

-0.15

GI

-0.15

own

-0.14

non

-0.14

za

-0.14

alat

-0.14

eman

-0.14

odelist

-0.14

POSITIVE LOGITS

 then

0.30

then

0.27

åĪĻ

0.23

THEN

0.22

 else

0.22

 Then

0.21

Then

0.21

 THEN

0.20

 otherwise

0.19

 entonces

0.18

Activations Density 0.174%

colons followed by non-zero values

No Comments

No Known Activations