INDEX

Explanations

non-English characters or symbols, particularly in a specific context

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

estatus

-0.19

aidu

-0.17

abbo

-0.16

@brief

-0.16

vise

-0.16

avou

-0.16

mrt

-0.16

æĸĻçĦ¡æĸĻ

-0.15

byss

-0.15

vail

-0.15

POSITIVE LOGITS

è¼ī

0.14

Dod

0.14

ill

0.13

Mou

0.13

 Mara

0.13

ëģĶ

0.13

 Enlightenment

0.13

ãģĭãģ£ãģ¦

0.13

::*

0.13

ILL

0.13

Activations Density 0.003%

non-English characters or symbols, particularly in a specific context

No Comments

No Known Activations