INDEX

Explanations

phrases indicating confusion or ambiguity regarding ownership or responsibilities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 22-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L22R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.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

avian

-0.20

adil

-0.15

etsk

-0.15

_SIG

-0.14

,:,

-0.14

 Trot

-0.14

ano

-0.14

istrat

-0.13

aceutical

-0.13

aven

-0.13

POSITIVE LOGITS

 nothing

0.40

nothing

0.38

 Nothing

0.36

Nothing

0.36

 NOTHING

0.35

 nada

0.31

 nichts

0.29

 anything

0.28

anything

0.26

 rien

0.22

Activations Density 0.064%

phrases indicating confusion or ambiguity regarding ownership or responsibilities

No Comments

No Known Activations