INDEX

Explanations

significant actions or states that have been recognized or acknowledged by others

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

.scalablytyped

-0.15

 however

-0.15

itu

-0.14

rst

-0.14

sez

-0.13

aghan

-0.13

anske

-0.13

arness

-0.13

rst

-0.13

POSITIVE LOGITS

 accordingly

0.16

è³¢

0.16

its

0.15

ewise

0.14

 convolution

0.14

roadcast

0.14

_FATAL

0.14

 consequently

0.14

Its

0.14

iesz

0.14

Activations Density 0.611%

significant actions or states that have been recognized or acknowledged by others

No Comments

No Known Activations