INDEX

Explanations

references to size, particularly large and massive quantities or entities

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

Negative Logits

 little

-0.19

 smaller

-0.18

çļĦå°ı

-0.18

 tiny

-0.18

 small

-0.18

 lille

-0.17

 slight

-0.16

 pequ

-0.16

 smallest

-0.16

ouro

-0.16

POSITIVE LOGITS

-scale

0.56

 scale

0.33

scale

0.31

-format

0.30

 Scale

0.28

Scale

0.28

_scale

0.25

 enough

0.24

-hearted

0.24

 amounts

0.24

Activations Density 0.054%

references to size, particularly large and massive quantities or entities

No Known Activations