INDEX

Explanations

references to political events and sports outcomes

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

eton

-0.14

ncy

-0.14

bone

-0.14

omal

-0.13

uida

-0.13

raid

-0.13

 dÆ°á»£c

-0.13

mmc

-0.13

.pres

-0.13

uria

-0.13

POSITIVE LOGITS

 score

0.18

 ÑĢÐ°Ñħ

0.18

nil

0.18

ware

0.18

 Score

0.17

 ÑĢÐ°ÑħÑĥÐ½Ð¾Ðº

0.17

WARE

0.17

-nil

0.16

 ÑģÑĩÐµÑĤ

0.16

rames

0.15

Activations Density 0.139%

references to political events and sports outcomes

No Comments

No Known Activations