INDEX

Explanations

references to food and healthy eating

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

umb

-0.15

hog

-0.15

 gravy

-0.14

 gravity

-0.14

ë¬

-0.14

Beer

-0.14

éĨ´

-0.14

ÑĥÐ¶Ðµ

-0.14

 shortest

-0.14

 potatoes

-0.13

POSITIVE LOGITS

 healthy

0.29

 health

0.27

 HEALTH

0.26

 guilt

0.26

åģ¥åº·

0.25

healthy

0.25

 healthier

0.24

health

0.24

 Healthy

0.24

 fiber

0.23

Activations Density 0.093%

references to food and healthy eating

No Comments

No Known Activations