INDEX

Explanations

terms related to long-duration concepts or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

ikal

-0.15

ipe

-0.15

Ð¸Ð¿

-0.15

NAV

-0.15

NAS

-0.15

/***/

-0.15

ozo

-0.15

iping

-0.15

 addCriterion

-0.15

woff

-0.15

POSITIVE LOGITS

Ã©n

0.16

agate

0.16

enek

0.16

adera

0.15

Ð´Ðµ

0.14

ourt

0.14

aliz

0.14

åĵ

0.14

Activations Density 0.024%

terms related to long-duration concepts or entities

No Comments

No Known Activations