INDEX

Explanations

vocabulary related to gameplay mechanics and dice rolling

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

èĶµ

-0.15

duct

-0.14

 Mour

-0.14

 brush

-0.14

Ã¡ty

-0.14

fik

-0.14

Ãºs

-0.14

intree

-0.13

 Brush

-0.13

æĸĹ

-0.13

POSITIVE LOGITS

 rolls

0.45

 roll

0.44

 rolling

0.42

-roll

0.41

 dice

0.40

 rolled

0.39

 Rolling

0.38

 Roll

0.38

roll

0.38

Roll

0.36

Activations Density 0.029%

vocabulary related to gameplay mechanics and dice rolling

No Comments

No Known Activations