INDEX

Explanations

references to components in various contexts

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

ough

-0.20

ãģ¾ãģŁ

-0.17

orie

-0.16

ups

-0.15

iben

-0.15

orta

-0.15

speaker

-0.15

enga

-0.15

boys

-0.15

arp

-0.15

POSITIVE LOGITS

ized

0.20

æĿĲ

0.18

alist

0.18

\HttpFoundation

0.17

errupted

0.15

 NotSupportedException

0.15

0.14

uition

0.14

iert

0.14

Activations Density 0.100%

references to components in various contexts

No Comments

No Known Activations