INDEX

Explanations

phrases indicating ranking or ordinal relationships

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

conde

-0.15

voj

-0.14

anches

-0.14

prises

-0.14

ield

-0.14

nett

-0.14

oints

-0.14

quine

-0.13

bler

-0.13

deviceId

-0.13

POSITIVE LOGITS

 them

0.23

such

0.22

 such

0.22

 Such

0.18

 SUCH

0.18

è¿Ļæł·çļĦ

0.17

Such

0.17

 Ð½Ð¸Ñħ

0.17

them

0.17

 suche

0.16

Activations Density 0.053%

phrases indicating ranking or ordinal relationships

No Comments

No Known Activations