INDEX

Explanations

the word "Do" as an imperative or question prompt

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.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

uyo

-0.19

ipi

-0.17

usto

-0.16

.tv

-0.15

nnen

-0.15

anco

-0.14

ungan

-0.14

uy

-0.14

'gc

-0.14

pi

-0.13

POSITIVE LOGITS

zens

0.20

antes

0.17

anter

0.16

seg

0.15

SENT

0.15

orm

0.14

olie

0.14

ÑĥÑĩÐ°

0.14

zen

0.14

 tang

0.14

Activations Density 0.029%

the word "Do" as an imperative or question prompt

No Comments

No Known Activations