INDEX

Explanations

punctuation marks and formatting elements commonly found in documentation or articles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

onn

-0.17

onne

-0.16

wen

-0.15

allo

-0.15

ollo

-0.15

antal

-0.15

uiltin

-0.14

ardy

-0.14

 Tubes

-0.14

adan

-0.14

POSITIVE LOGITS

osate

0.15

resh

0.15

iban

0.14

plr

0.14

eterminate

0.14

apper

0.14

abama

0.14

ì²ł

0.14

SPATH

0.14

ActionCreators

0.13

Activations Density 0.004%

punctuation marks and formatting elements commonly found in documentation or articles

No Comments

No Known Activations