INDEX

Explanations

references to historical landmarks and preservation efforts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

ois

-0.15

useum

-0.15

seo

-0.14

opo

-0.14

notated

-0.14

woord

-0.14

OU

-0.14

egen

-0.14

éŁ¿

-0.14

otype

-0.14

POSITIVE LOGITS

 Registers

0.17

NR

0.17

 Architect

0.17

_COMPAT

0.16

 Register

0.16

registers

0.16

Pk

0.16

wij

0.16

 Historic

0.16

 Criteria

0.16

Activations Density 0.011%

references to historical landmarks and preservation efforts

No Comments

No Known Activations