INDEX

Explanations

references to manipulation and specific names or terms associated with communication

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.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

ally

-0.16

emm

-0.16

_malloc

-0.16

phant

-0.16

 Shame

-0.15

Ð¹

-0.15

OrNil

-0.14

FromArray

-0.14

çŁ¢

-0.14

ablish

-0.14

POSITIVE LOGITS

tras

0.23

uales

0.20

resa

0.20

uela

0.20

uring

0.19

hattan

0.19

ulative

0.19

ifold

0.18

uelle

0.18

raq

0.18

Activations Density 0.032%

references to manipulation and specific names or terms associated with communication

No Comments

No Known Activations