INDEX

Explanations

prepositions indicating location or context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

arrow

-0.16

aines

-0.15

heck

-0.15

icon

-0.14

å¥Ī

-0.14

ora

-0.14

_itr

-0.14

CAPE

-0.14

èĮĥ

-0.13

èĿ

-0.13

POSITIVE LOGITS

 action

0.32

 Action

0.28

-action

0.25

 ACTION

0.25

Action

0.22

_action

0.22

.action

0.21

 Ð´ÑĸÑĹ

0.21

action

0.20

/action

0.20

Activations Density 0.181%

prepositions indicating location or context

No Comments

No Known Activations