INDEX

Explanations

mathematical notation and formal expressions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

rome

-0.17

imde

-0.17

ãģ£ãģ¨

-0.16

plusplus

-0.16

 unders

-0.15

acent

-0.15

Ä±rÄ±

-0.14

 Rounds

-0.14

 Council

-0.14

irect

-0.14

POSITIVE LOGITS

vern

0.17

herit

0.15

VERN

0.15

ÑĢÑĥÑĩ

0.14

udad

0.14

velt

0.14

0.13

 esac

0.13

AMP

0.13

_locked

0.13

Activations Density 0.022%

mathematical notation and formal expressions

No Comments

No Known Activations