INDEX

Explanations

references to behavioral traits and concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

ye

-0.21

yan

-0.16

eye

-0.15

ars

-0.15

attended

-0.15

ä½Ļ

-0.15

egal

-0.15

sburgh

-0.15

ya

-0.15

POSITIVE LOGITS

ally

0.31

ial

0.27

 patterns

0.22

 modification

0.20

istically

0.19

isms

0.19

patterns

0.19

IAL

0.19

alist

0.19

 Modification

0.19

Activations Density 0.023%

references to behavioral traits and concepts

No Comments

No Known Activations