INDEX

Explanations

references to "Kon" or related terms, suggesting a focus on content related to structure or organization

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.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

-0.15

 Undert

-0.14

 Austral

-0.14

-www

-0.14

alling

-0.14

andas

-0.13

 Ekon

-0.13

 finally

-0.13

å¹¸

-0.13

 Detect

-0.13

POSITIVE LOGITS

linger

0.19

ymous

0.18

rad

0.16

ederland

0.15

RLF

0.15

nect

0.15

enate

0.15

exion

0.15

ildo

0.14

 zbyt

0.14

Activations Density 0.014%

references to "Kon" or related terms, suggesting a focus on content related to structure or organization

No Comments

No Known Activations