INDEX

Explanations

place names and educational institutions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

ãĥ¼ãĥª

-0.19

avia

-0.19

erea

-0.17

ocup

-0.16

avÄĽ

-0.15

 finger

-0.15

Ð¶Ð¸

-0.15

ya

-0.15

acades

-0.15

iel

-0.15

POSITIVE LOGITS

 Sanat

0.25

Rad

0.21

 Media

0.21

 Ridley

0.21

 Upper

0.20

Rad

0.20

Gul

0.20

Nor

0.19

 Chester

0.19

Ard

0.18

Activations Density 0.016%

place names and educational institutions

No Comments

No Known Activations