INDEX

Explanations

the word "i" and the word "och" which typically relate to inclusiveness or connection

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

(~(

-0.17

illac

-0.15

roker

-0.15

;element

-0.15

agon

-0.14

rosse

-0.14

imler

-0.14

 Bord

-0.13

 Geld

-0.13

unya

-0.13

POSITIVE LOGITS

rub

0.15

 Rubio

0.15

è±Ĩ

0.14

AINS

0.14

Rub

0.14

rew

0.14

 Vine

0.13

 inline

0.13

anner

0.13

 overst

0.13

Activations Density 0.002%

the word "i" and the word "och" which typically relate to inclusiveness or connection

No Comments

No Known Activations