INDEX

Explanations

comparisons and choices regarding preferences or selections

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

Tro

-0.16

tring

-0.15

udes

-0.15

 Medina

-0.14

Th

-0.14

Wol

-0.14

 Feld

-0.13

azon

-0.13

fe

-0.13

POSITIVE LOGITS

_Component

0.16

anus

0.15

abyrin

0.15

cran

0.15

 beide

0.15

Ø§ÙĦØª

0.15

:both

0.15

aÃ±

0.15

ë°

0.15

ilden

0.15

Activations Density 0.264%

comparisons and choices regarding preferences or selections

No Comments

No Known Activations