INDEX

Explanations

occurrences of news-related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

 cÃ´t

-0.16

 Giuliani

-0.15

è«

-0.14

ãĤ·ãĥ¼

-0.14

asco

-0.14

.construct

-0.14

ites

-0.14

 cycle

-0.14

erv

-0.13

erville

-0.13

POSITIVE LOGITS

 news

0.21

etat

0.18

abyrin

0.16

ãĥ¼ãĤ¹

0.16

ulen

0.15

 developments

0.15

 updates

0.15

flash

0.14

 LÃ¢m

0.14

ÄĽn

0.14

Activations Density 0.083%

occurrences of news-related terms

No Comments

No Known Activations