INDEX

Explanations

references to names or terms related to "Lor."

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

isper

-0.16

pts

-0.15

åĬ¨çĶŁæĪĲ

-0.15

landing

-0.15

VOID

-0.15

ect

-0.15

mium

-0.15

heits

-0.15

::$_

-0.14

vironments

-0.14

POSITIVE LOGITS

 Nose

0.18

 Ipsum

0.17

hei

0.16

Lor

0.16

@if

0.15

omore

0.15

 nose

0.15

lub

0.14

 ipsum

0.14

ochen

0.14

Activations Density 0.017%

references to names or terms related to "Lor."

No Comments

No Known Activations