INDEX

Explanations

expressions of good wishes and encouragement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

zar

-0.14

leston

-0.14

hev

-0.14

endor

-0.14

EventHandler

-0.14

ndo

-0.14

oles

-0.14

Ð¾ÑģÑĤÐ°ÑĤ

-0.13

utzer

-0.13

POSITIVE LOGITS

 wherever

0.20

 wishes

0.19

 everyone

0.17

 everybody

0.16

everyone

0.15

 indeed

0.15

edis

0.15

vir

0.15

leep

0.14

ãĥ³ãĤº

0.14

Activations Density 0.046%

expressions of good wishes and encouragement

No Comments

No Known Activations