INDEX

Explanations

the term "troll" and its variations, indicating discussions related to internet trolling behavior

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.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

sort

-0.16

ÙĦØ§Ùħ

-0.15

ience

-0.14

 Kern

-0.14

iente

-0.14

 Plans

-0.14

 racket

-0.14

emouth

-0.14

idd

-0.14

arian

-0.14

POSITIVE LOGITS

auge

0.15

vester

0.14

alten

0.14

.Css

0.14

adic

0.14

ulas

0.13

Ø¨Ø§ÙĦ

0.13

Ð¾Ð¸ÑĤ

0.13

akeup

0.13

Ã¸d

0.13

Activations Density 0.005%

the term "troll" and its variations, indicating discussions related to internet trolling behavior

No Comments

No Known Activations