INDEX

Explanations

proper nouns and names, particularly those relating to people and organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

obuf

-0.15

chant

-0.15

slow

-0.15

ariat

-0.15

ãĥ³ãĥĦ

-0.15

.sharedInstance

-0.15

æģµ

-0.14

inyin

-0.14

lexer

-0.14

uisse

-0.14

POSITIVE LOGITS

azon

0.19

izon

0.17

izons

0.17

ilitation

0.16

icopt

0.16

icopter

0.15

zÃ¡

0.15

oldt

0.15

orthand

0.15

.uml

0.15

Activations Density 0.112%

proper nouns and names, particularly those relating to people and organizations

No Comments

No Known Activations