INDEX

Explanations

instances of the word "pretty" and its variations, indicating a search for expressions of mild emphasis or positive sentiment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

istrat

-0.16

 overly

-0.15

st

-0.15

ulumi

-0.15

icerca

-0.15

mys

-0.14

lot

-0.14

 Difficulty

-0.14

roma

-0.14

hma

-0.14

POSITIVE LOGITS

 much

0.30

-ÑĤÐ°ÐºÐ¸

0.26

 darn

0.23

much

0.22

 Much

0.21

-boy

0.21

Much

0.19

 MUCH

0.19

assin

0.18

Activations Density 0.016%

instances of the word "pretty" and its variations, indicating a search for expressions of mild emphasis or positive sentiment

No Comments

No Known Activations