INDEX

Explanations

phrases that denote past actions and experiences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

Cao

-0.14

.SIZE

-0.14

-Line

-0.14

aldo

-0.14

fu

-0.14

insky

-0.13

ufen

-0.13

çĪ¾

-0.13

oppers

-0.13

Ãºi

-0.13

POSITIVE LOGITS

 yesterday

0.27

 last

0.26

 earlier

0.26

 recently

0.21

 initially

0.19

 Ø³Ø§Ø¨ÙĤ

0.18

 originally

0.18

last

0.18

 formerly

0.18

 during

0.17

Activations Density 0.415%

phrases that denote past actions and experiences

No Comments

No Known Activations