INDEX

Explanations

references to historical places and localities

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

wash

-0.17

ØºØ§ÙĦ

-0.17

à¹Ģà¸ª

-0.17

_LOGGER

-0.16

rored

-0.16

 Somerset

-0.15

çīĻ

-0.15

æ²Ī

-0.15

.scalablytyped

-0.15

SWG

-0.14

POSITIVE LOGITS

Hen

0.23

 Richmond

0.23

Rich

0.21

Hen

0.18

Din

0.17

nder

0.16

Mech

0.16

 ÐĴÐ°ÑĢ

0.16

 Petersburg

0.15

usan

0.15

Activations Density 0.012%

references to historical places and localities

No Comments

No Known Activations