INDEX

Explanations

references to historical events and military terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

¬ģ

-0.15

oÅ¾

-0.15

UBLE

-0.15

Ð¾ÑģÐ¿

-0.15

 Hoff

-0.14

resco

-0.14

angl

-0.14

cul

-0.14

etter

-0.13

ubs

-0.13

POSITIVE LOGITS

 Ø¹Ø´Ø±

0.19

 baseman

0.15

 division

0.15

 anniversary

0.15

-order

0.15

 timers

0.14

 phase

0.14

orWhere

0.14

arily

0.14

.wp

0.14

Activations Density 0.097%

references to historical events and military terminology

No Comments

No Known Activations