INDEX

Explanations

references to narrative techniques and film analysis

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

 juxtap

-0.16

primir

-0.16

 existence

-0.15

zeich

-0.15

Rt

-0.15

upp

-0.14

 vign

-0.14

 rendez

-0.14

 medicine

-0.14

 Scientists

-0.14

POSITIVE LOGITS

 Formal

0.34

 formal

0.32

 formally

0.24

 structural

0.23

 Structural

0.20

structural

0.20

 compos

0.19

Struct

0.18

itez

0.18

ORMAL

0.17

Activations Density 0.159%

references to narrative techniques and film analysis

No Comments

No Known Activations