INDEX

Explanations

terms related to precision or specific measurements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

herits

-0.17

braco

-0.17

ÅĻ

-0.15

 Cooler

-0.15

acic

-0.15

flater

-0.15

ochond

-0.15

 limburg

-0.15

FINITY

-0.14

paged

-0.14

POSITIVE LOGITS

 prec

0.37

 Prec

0.37

prec

0.36

Prec

0.33

ise

0.30

 precip

0.26

ip

0.25

ursors

0.25

 prÃ©c

0.24

inct

0.23

Activations Density 0.012%

terms related to precision or specific measurements

No Comments

No Known Activations