INDEX

Explanations

keywords related to explanations, discussions, and comparisons within educational and analytical contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

 &&↵

-0.17

ãģĿãģĹãģ¦

-0.16

AndGet

-0.15

—and

-0.15

 &&č↵

-0.14

áº¡t

-0.13

&&!

-0.13

&&

-0.13

&action

-0.13

âĢ¦and

-0.13

POSITIVE LOGITS

or

0.96

 Ð¸Ð»Ð¸

0.75

æĪĸèĢħ

0.73

 æĪĸ

0.72

 oder

0.70

æĪĸ

0.69

 atau

0.67

 nebo

0.66

 hoáº·c

0.66

 ÛĮØ§

0.63

Activations Density 1.491%

keywords related to explanations, discussions, and comparisons within educational and analytical contexts

No Comments

No Known Activations