INDEX

Explanations

elements related to branding and design symbolism

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

å¦ĥ

-0.17

.AutoComplete

-0.16

ãģ¡ãĤĥãĤĵ

-0.15

remen

-0.14

avier

-0.14

 Ø®Ø§ÙĨ

-0.14

Ð»ÑĥÐ³

-0.14

adan

-0.14

isay

-0.14

rien

-0.14

POSITIVE LOGITS

ãĥ«ãĥī

0.21

cla

0.17

 Ø§ÙĦØ§Ø¹

0.15

Ĥæķ°

0.15

ãĥªãĤ«

0.14

.Xaml

0.14

lac

0.14

 frac

0.14

DUCT

0.13

(typ

0.13

Activations Density 0.064%

elements related to branding and design symbolism

No Comments

No Known Activations