INDEX

Explanations

references to developers and their interactions with software or tools

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

 almost

-0.19

Almost

-0.17

almost

-0.17

 Almost

-0.16

å¹¾

-0.16

bject

-0.16

 neredeyse

-0.16

dech

-0.15

 Says

-0.15

ragon

-0.14

POSITIVE LOGITS

 somehow

0.40

 perhaps

0.34

 maybe

0.30

perhaps

0.29

 somewhere

0.29

maybe

0.26

 Perhaps

0.25

 Ð»Ð¸Ð±Ð¾

0.25

Perhaps

0.24

or

0.23

Activations Density 0.424%

references to developers and their interactions with software or tools

No Comments

No Known Activations