INDEX

Explanations

occurrences of the word "in."

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

 Coord

-0.15

LOPT

-0.15

evi

-0.14

formation

-0.14

 fate

-0.14

 Regards

-0.14

ffects

-0.14

ç¡

-0.13

ÑģÐ»Ð¸

-0.13

illin

-0.13

POSITIVE LOGITS

won

0.21

midd

0.20

wend

0.18

der

0.18

enting

0.17

zag

0.17

het

0.17

 Europa

0.16

.cx

0.16

span

0.15

Activations Density 0.008%

occurrences of the word "in."

No Comments

No Known Activations