INDEX

Explanations

references to architectural features and improvements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

 inside

-0.18

Occurrences

-0.17

inside

-0.17

 Inside

-0.16

Ã©o

-0.15

endencies

-0.15

 fork

-0.15

 Fork

-0.14

Inside

-0.14

æ³Ľ

-0.14

POSITIVE LOGITS

 exterior

0.18

External

0.16

 ìĻ¸ë¶Ģ

0.16

 external

0.16

å¤ĸ

0.15

 External

0.15

 outdoor

0.15

.getExternal

0.15

 å¤ĸ

0.15

 Ð³ÑĢÐ°Ð´Ñĥ

0.15

Activations Density 0.089%

references to architectural features and improvements

No Comments

No Known Activations