INDEX

Explanations

occurrences of the word "do" and its variations, indicating emphasis on actions or commands

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

maids

-0.17

loi

-0.16

cular

-0.16

ë²½

-0.16

pery

-0.15

ditor

-0.15

stoup

-0.15

giene

-0.15

riba

-0.15

OrDefault

-0.14

POSITIVE LOGITS

zens

0.22

/do

0.17

mund

0.15

cket

0.14

 Karn

0.14

getManager

0.14

Roc

0.14

age

0.14

tures

0.14

SENT

0.14

Activations Density 0.052%

occurrences of the word "do" and its variations, indicating emphasis on actions or commands

No Comments

No Known Activations