INDEX

Explanations

negations and the absence of certain actions or qualities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

uya

-0.15

urent

-0.14

inx

-0.14

Ð»Ð¾Ð´

-0.14

qus

-0.14

ero

-0.13

eniable

-0.13

 indemn

-0.13

uby

-0.13

oda

-0.13

POSITIVE LOGITS

Ã¨ve

0.16

 sonst

0.15

apas

0.15

 other

0.15

 DISP

0.15

try

0.14

-Ta

0.14

à¹Ģà¸¥à¸¢

0.14

oodle

0.14

exels

0.14

Activations Density 0.218%

negations and the absence of certain actions or qualities

No Comments

No Known Activations