INDEX

Explanations

proper nouns, particularly names of individuals and entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

oenix

-0.14

 Press

-0.14

ongan

-0.14

tra

-0.13

andas

-0.13

press

-0.13

 colors

-0.13

åı«

-0.13

Ã¡d

-0.13

FORCE

-0.12

POSITIVE LOGITS

 ÙĩÙħÚĨÙĨÛĮÙĨ

0.17

odore

0.17

akedown

0.15

 ayrÄ±ca

0.14

zbollah

0.14

adors

0.13

 TMPro

0.13

Ø¹Ø§ÙĦ

0.13

ëĤľ

0.13

_ops

0.13

Activations Density 0.040%

proper nouns, particularly names of individuals and entities

No Comments

No Known Activations