INDEX

Explanations

terms related to decision-making processes in a network

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

æĬĺ

-0.16

 NhÃ¢n

-0.15

oref

-0.15

enco

-0.14

emann

-0.14

tran

-0.14

 Harmon

-0.14

 automatic

-0.14

 Hunt

-0.13

rab

-0.13

POSITIVE LOGITS

 oracle

0.19

 Nielsen

0.17

ffffffff

0.17

uisse

0.17

 Alice

0.16

 Corruption

0.15

oracle

0.15

 PartialView

0.15

 Challenger

0.15

 Commit

0.15

Activations Density 0.019%

terms related to decision-making processes in a network

No Comments

No Known Activations