INDEX

Explanations

references to professional wrestling events and matches

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

.managed

-0.18

Reusable

-0.18

elay

-0.17

zug

-0.15

agli

-0.15

ooth

-0.14

 McCart

-0.14

Ø³Ø§Øª

-0.14

 schl

-0.14

akte

-0.14

POSITIVE LOGITS

NJ

0.20

AE

0.19

 Impact

0.18

Ä±rak

0.18

nty

0.18

zym

0.17

 Bullet

0.17

 Ward

0.17

PW

0.16

 Kenny

0.16

Activations Density 0.036%

references to professional wrestling events and matches

No Comments

No Known Activations