INDEX

Explanations

phrases associated with good and bad news

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

Not in Any Lists

No Comments

Negative Logits

uhn

-0.14

inz

-0.14

à¤¾à¤ªà¤¨

-0.14

ÄĽk

-0.14

inin

-0.14

indsight

-0.14

.AddParameter

-0.13

erialized

-0.13

istributed

-0.13

 æĦ

-0.13

POSITIVE LOGITS

 news

1.13

news

0.92

 News

0.85

 NEWS

0.81

News

0.79

-news

0.79

_news

0.69

NEWS

0.67

(news

0.66

/news

0.63

Activations Density 0.292%

phrases associated with good and bad news

No Comments

No Known Activations