INDEX

Explanations

references to specific genres or types of media

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 russian

-0.20

 canadian

-0.19

æĪ¶

-0.17

 praÅ¾

-0.15

 DÄĽ

-0.15

isson

-0.15

 russ

-0.15

 Irish

-0.15

 paris

-0.15

 Croatian

-0.15

POSITIVE LOGITS

Rum

0.28

 Indie

0.24

USA

0.18

Ind

0.18

 Ð¡Ð¨ÐĲ

0.18

Jug

0.17

 Ð¡Ð¡Ð¡Ðł

0.17

Kub

0.17

 Core

0.17

Activations Density 0.057%

references to specific genres or types of media

No Comments

No Known Activations