INDEX

Explanations

comments in code

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

áº¡o

-0.15

ewire

-0.14

äºİ

-0.14

ior

-0.14

iyatÄ±

-0.14

 Owens

-0.14

glas

-0.14

inerary

-0.13

eld

-0.13

hlas

-0.13

POSITIVE LOGITS

 noqa

0.25

<--

0.24

<-

0.20

<-

0.18

oret

0.16

sic

0.16

eslint

0.16

icina

0.16

 ìĸĺ

0.15

NOI

0.15

Activations Density 0.024%

comments in code

No Comments

No Known Activations