INDEX

Explanations

specific numerical or coded values, particularly in technical and statistical contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

 defence

-0.19

 Spanish

-0.18

 proposals

-0.17

Spanish

-0.17

Spain

-0.17

 proposal

-0.17

proposal

-0.16

 Madrid

-0.16

 spanish

-0.15

 honour

-0.15

POSITIVE LOGITS

 Belize

0.16

 reported

0.15

 Panama

0.15

 registr

0.15

 inform

0.15

 informed

0.14

 registered

0.14

 reporting

0.14

 informing

0.14

 registration

0.14

Activations Density 0.004%

specific numerical or coded values, particularly in technical and statistical contexts

No Comments

No Known Activations