INDEX

Explanations

words associated with relationships and relevance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

Relations

-0.18

rah

-0.17

ses

-0.17

_relationship

-0.16

rial

-0.16

 Relations

-0.16

_relations

-0.16

fall

-0.15

istrat

-0.15

ernet

-0.15

POSITIVE LOGITS

ness

0.25

èģĶ

0.19

-party

0.19

ly

0.19

adele

0.17

icut

0.17

SHIP

0.17

 matters

0.16

ëĲľ

0.16

 matter

0.15

Activations Density 0.054%

words associated with relationships and relevance

No Comments

No Known Activations