INDEX

Explanations

references to unresolved story elements and endings in narratives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

 kÃ¶ÅŁ

-0.14

hek

-0.14

erspective

-0.14

æ¡Ĥ

-0.14

Î¼Î¿Î½

-0.14

 opak

-0.13

utsch

-0.13

ç°

-0.13

bish

-0.13

TestId

-0.13

POSITIVE LOGITS

 leave

0.29

 leaving

0.29

 leaves

0.27

 left

0.27

 cliff

0.26

 ending

0.25

 Leave

0.24

leave

0.23

 Leaving

0.22

Leave

0.22

Activations Density 0.108%

references to unresolved story elements and endings in narratives

No Comments

No Known Activations