INDEX

Explanations

phrases related to news and updates from various sources

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

ÏĦÏĥ

-0.19

Æ°á»£ng

-0.15

 sÃ¶

-0.14

åŃ£

-0.14

anik

-0.14

zych

-0.14

blr

-0.13

apons

-0.13

stown

-0.13

 pageInfo

-0.13

POSITIVE LOGITS

 news

0.51

 breaking

0.44

 News

0.44

news

0.41

 Breaking

0.41

 headlines

0.40

-news

0.39

 NEWS

0.39

Breaking

0.38

News

0.37

Activations Density 0.180%

phrases related to news and updates from various sources

No Comments

No Known Activations