INDEX

Explanations

phrases or patterns that suggest possession or attribution

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

ongan

-0.17

sandbox

-0.16

spar

-0.15

illez

-0.15

loquent

-0.15

elerik

-0.15

çļĦä¸Ģä¸ª

-0.15

erence

-0.14

ittest

-0.14

acades

-0.14

POSITIVE LOGITS

ones

0.18

vak

0.15

place

0.14

-times

0.14

aran

0.14

ogg

0.14

PACE

0.14

 audible

0.13

sonian

0.13

may

0.13

Activations Density 0.033%

phrases or patterns that suggest possession or attribution

No Comments

No Known Activations