INDEX

Explanations

references to architectural elements and projects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

anÃ¡

-0.15

åı

-0.14

 gravity

-0.14

Ð¸ÑĨ

-0.14

à¹Īà¸³

-0.14

èĪĮ

-0.14

COPE

-0.14

kop

-0.13

gravity

-0.13

ussen

-0.13

POSITIVE LOGITS

 bibli

0.20

 Bibli

0.19

ibli

0.17

 balk

0.17

 stall

0.17

 galer

0.17

 multif

0.17

 apart

0.17

WC

0.17

 appart

0.17

Activations Density 0.205%

references to architectural elements and projects

No Comments

No Known Activations