INDEX

Explanations

references to the concept of normalcy

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

eler

-0.17

ern

-0.17

eling

-0.17

ary

-0.16

inerary

-0.15

ernet

-0.15

esh

-0.15

emic

-0.15

.GroupLayout

-0.15

essel

-0.14

POSITIVE LOGITS

mente

0.33

ity

0.24

ised

0.23

-sized

0.22

cy

0.22

izr

0.20

izedName

0.20

izing

0.19

isation

0.19

-looking

0.19

Activations Density 0.040%

references to the concept of normalcy

No Comments

No Known Activations