INDEX

Explanations

actions or states that indicate evaluation or classification

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

-as

-0.28

As

-0.26

As

-0.25

.As

-0.22

_AS

-0.21

_as

-0.21

_As

-0.20

/as

-0.20

 Ashton

-0.18

AS

-0.17

POSITIVE LOGITS

 merely

0.17

è§

0.16

 altern

0.16

Ø²Ø§Ø±

0.16

icky

0.15

ÃŃna

0.15

adh

0.14

perPage

0.14

urs

0.14

ëĶĶìĭľ

0.14

Activations Density 0.105%

actions or states that indicate evaluation or classification

No Comments

No Known Activations