INDEX

Explanations

specific identifiers or names related to titles and characters in a media context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

 zdrav

-0.17

 geschichten

-0.16

 functional

-0.15

//===

-0.15

undy

-0.15

 Functional

-0.15

Ã¤l

-0.14

YYS

-0.13

pct

-0.13

ingleton

-0.13

POSITIVE LOGITS

az

0.29

egy

0.28

sz

0.28

 Ã©s

0.26

 vagy

0.24

 harm

0.23

cs

0.23

 volt

0.23

 szer

0.23

Sz

0.22

Activations Density 0.004%

specific identifiers or names related to titles and characters in a media context

No Comments

No Known Activations