INDEX

Explanations

spiritual or religious language

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

ecast

-0.17

bral

-0.16

 Acres

-0.16

aph

-0.15

Ã¯

-0.15

awah

-0.15

onus

-0.15

.Automation

-0.15

ÑĪÐ¸Ð±

-0.15

erie

-0.14

POSITIVE LOGITS

.rs

0.21

Ðµ

0.20

cca

0.20

 Serbian

0.19

 Kral

0.19

 ÄĲ

0.19

 Bosnia

0.19

Bos

0.19

Voj

0.18

 Äĳ

0.18

Activations Density 0.106%

spiritual or religious language

No Comments

No Known Activations