INDEX

Explanations

references to "pictures" or "visual imagery."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

Ùĳ

-0.15

yla

-0.15

esin

-0.15

ìį¨

-0.15

ra

-0.15

wn

-0.15

 Cheer

-0.14

ëĭ¤ê°Ģ

-0.14

ipay

-0.14

ilet

-0.14

POSITIVE LOGITS

-per

0.24

orial

0.21

 perfect

0.20

perfect

0.20

que

0.20

ocks

0.19

Perfect

0.19

ofday

0.19

/video

0.18

 Perfect

0.17

Activations Density 0.027%

references to "pictures" or "visual imagery."

No Comments

No Known Activations