INDEX

Explanations

terms related to hierarchy and superiority

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 27-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L27R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.27.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

afari

-0.19

sworth

-0.16

suppress

-0.15

nest

-0.15

/how

-0.14

voor

-0.14

pray

-0.14

ACC

-0.14

doors

-0.14

soever

-0.14

POSITIVE LOGITS

ior

0.28

iors

0.27

intendent

0.24

IOR

0.24

iore

0.23

charged

0.23

cil

0.23

ordinate

0.22

stit

0.22

iores

0.21

Activations Density 0.039%

terms related to hierarchy and superiority

No Comments

No Known Activations