INDEX

Explanations

definite articles and their occurrences in context

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

 opportunity

-0.18

 likes

-0.17

 entire

-0.16

 stuff

-0.15

atts

-0.15

 avenue

-0.15

 likeness

-0.15

äººæīį

-0.15

äºĪ

-0.14

 Entire

-0.14

POSITIVE LOGITS

few

0.44

few

0.36

 many

0.33

Few

0.33

Few

0.31

many

0.30

åĩłä¸ª

0.27

 several

0.25

 rare

0.24

 MANY

0.24

Activations Density 0.117%

definite articles and their occurrences in context

No Comments

No Known Activations