INDEX

Explanations

punctuation marks, particularly periods and commas, indicating the end of sentences and clauses

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

oran

-0.15

 alike

-0.14

ÐĴÐ¡

-0.14

 saja

-0.14

dsa

-0.14

 Tanner

-0.13

.synthetic

-0.13

Ã¼lÃ¼

-0.13

etti

-0.13

igne

-0.13

POSITIVE LOGITS

Fol

0.24

Ir

0.23

 Right

0.22

 Bear

0.21

 Princip

0.20

 Bene

0.20

Mer

0.20

 Listed

0.20

Prim

0.20

 Prim

0.20

Activations Density 0.041%

punctuation marks, particularly periods and commas, indicating the end of sentences and clauses

No Comments

No Known Activations