INDEX

Explanations

themes related to ugliness and imperfections

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 sine

-0.16

riel

-0.15

lake

-0.15

 propName

-0.15

ev

-0.14

undred

-0.13

 ÙĤÙĦ

-0.13

ov

-0.13

ementia

-0.13

POSITIVE LOGITS

cly

0.16

ioc

0.15

aad

0.15

enstein

0.15

inspace

0.15

iddet

0.15

 Parkway

0.14

ipv

0.14

iets

0.14

cio

0.14

Activations Density 0.016%

themes related to ugliness and imperfections

No Comments

No Known Activations