INDEX

Explanations

specific non-standard characters or symbols

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

-0.17

¶ģ

-0.15

plier

-0.15

ereco

-0.14

ki

-0.14

eref

-0.14

ooth

-0.14

¦¬

-0.14

erman

-0.14

POSITIVE LOGITS

uong

0.16

ERGE

0.14

ÏĦÎ¹Î¿

0.14

ī

0.14

OWL

0.14

 Benton

0.13

GridColumn

0.13

bow

0.13

à¤°à¤£

0.13

nen

0.13

Activations Density 0.016%

specific non-standard characters or symbols

No Comments

No Known Activations