INDEX

Explanations

technical specifications and data structures

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

ike

-0.16

agna

-0.15

aze

-0.14

lev

-0.14

ject

-0.14

vac

-0.13

 duck

-0.13

aland

-0.13

rv

-0.13

POSITIVE LOGITS

Å¡et

0.15

rikes

0.15

ãĥ¼ãĥł

0.15

inded

0.15

iddet

0.14

YYS

0.14

arp

0.14

.Xaml

0.13

riott

0.13

Activations Density 0.189%

technical specifications and data structures

No Comments

No Known Activations