INDEX

Explanations

references to geographical locations and capitals

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

arro

-0.15

pn

-0.14

hood

-0.14

.aggregate

-0.14

INGLE

-0.14

ands

-0.14

tura

-0.14

edReader

-0.13

 Sabha

-0.13

 Ãĸn

-0.13

POSITIVE LOGITS

ilen

0.14

 quÃ¡n

0.14

ibase

0.14

ifar

0.14

ħ§

0.14

Ð¸ÑĤÐ¸

0.14

isoft

0.13

itorio

0.13

vez

0.13

TextWriter

0.13

Activations Density 0.096%

references to geographical locations and capitals

No Comments

No Known Activations