INDEX

Explanations

phrases and topics that specify details about various subjects

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

izable

-0.17

fault

-0.16

argout

-0.16

_about

-0.15

nul

-0.14

à¸±à¸ļà¸¡

-0.14

ÙħÙĨØ¯

-0.14

about

-0.14

hlen

-0.14

isable

-0.14

POSITIVE LOGITS

how

0.21

/from

0.20

/to

0.19

-NLS

0.19

the

0.19

-face

0.19

/by

0.18

 what

0.17

 Ø·Ø±ÙĬÙĤ

0.17

/of

0.17

Activations Density 0.146%

phrases and topics that specify details about various subjects

No Comments

No Known Activations