INDEX

Explanations

references to the word "Blue" and its various contexts or combinations

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 27-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L27R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.27.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

rei

-0.19

nee

-0.17

rone

-0.16

lei

-0.16

rad

-0.15

aving

-0.15

ship

-0.15

rvine

-0.15

li

-0.15

teen

-0.15

POSITIVE LOGITS

prints

0.37

grass

0.33

berry

0.28

-collar

0.27

berries

0.26

bell

0.24

bird

0.23

jay

0.23

bon

0.22

printed

0.21

Activations Density 0.021%

references to the word "Blue" and its various contexts or combinations

No Known Activations