INDEX

Explanations

variations of the word "scale" and related terms

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

æłª

-0.15

Î´Î¿

-0.15

QC

-0.14

 actionTypes

-0.14

inement

-0.14

à¸¶à¸ģ

-0.14

ç·Ĵ

-0.14

clipse

-0.13

esk

-0.13

POSITIVE LOGITS

lop

0.21

lops

0.18

 Scal

0.17

 scal

0.17

=sc

0.17

abis

0.17

opus

0.17

jer

0.16

ibur

0.15

Sac

0.15

Activations Density 0.007%

variations of the word "scale" and related terms

No Comments

No Known Activations