INDEX

Explanations

negations and phrases indicating impossibility or lack of necessity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

.mixin

-0.15

.calls

-0.14

ockets

-0.14

ãģ¹ãģį

-0.14

uslim

-0.14

eri

-0.14

plements

-0.14

erdale

-0.13

Ãºsqueda

-0.13

 sink

-0.13

POSITIVE LOGITS

alic

0.15

aland

0.15

åĽ°

0.15

ÙĪÙĬ

0.14

allee

0.14

à¸Ľà¸£

0.14

Ø±Ø¨

0.14

å¼

0.14

elden

0.14

FER

0.14

Activations Density 0.082%

negations and phrases indicating impossibility or lack of necessity

No Comments

No Known Activations