INDEX

Explanations

numeric data or specific values presented in textual form

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

ermann

-0.17

oor

-0.15

tit

-0.15

bil

-0.15

quia

-0.15

tit

-0.15

pad

-0.15

Nap

-0.14

annel

-0.14

.pb

-0.14

POSITIVE LOGITS

pare

0.17

 Fold

0.14

iros

0.14

ÐĶÐļ

0.14

 Ã¤n

0.14

ãĥ³ãĥĨ

0.14

xFFFF

0.13

 kaps

0.13

ÄŁ

0.13

.cm

0.13

Activations Density 0.000%

numeric data or specific values presented in textual form

No Comments

No Known Activations