INDEX

Explanations

non-alphabetic characters and sequences related to digital content and formatting

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.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

angkan

-0.17

owie

-0.17

Boost

-0.15

ãģĻãĤĮãģ°

-0.15

.boost

-0.14

fty

-0.14

hist

-0.14

çĬ¶

-0.14

ovi

-0.14

_gradient

-0.14

POSITIVE LOGITS

asar

0.19

abei

0.17

ular

0.15

Mine

0.15

 danmark

0.15

iris

0.14

ajor

0.14

Pooling

0.14

Ľ

0.14

asa

0.13

Activations Density 0.003%

non-alphabetic characters and sequences related to digital content and formatting

No Comments

No Known Activations

non-alphabetic characters and sequences related to digital content and formatting

No Comments

No Known Activations