INDEX

Explanations

references to logical concepts and structures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.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

kening

-0.17

ê²

-0.17

argas

-0.15

zsche

-0.15

ego

-0.15

udeau

-0.14

etten

-0.14

lsen

-0.14

ushima

-0.14

 marg

-0.14

POSITIVE LOGITS

agma

0.17

olon

0.15

 Naming

0.14

ãĤ°ãĥ«

0.14

ved

0.14

ç´Ģ

0.14

Ø³ÙĪØ¨

0.14

oron

0.14

aces

0.13

 Sequential

0.13

Activations Density 0.049%

references to logical concepts and structures

No Comments

No Known Activations