INDEX

Explanations

unique numerical data and references related to scientific studies or methodologies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 27-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L27R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.27.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

lete

-0.15

Trad

-0.14

ilder

-0.14

landÄ±

-0.14

asu

-0.14

ÑĢÐ°Ð¿

-0.13

rente

-0.13

enia

-0.13

alculate

-0.12

akov

-0.12

POSITIVE LOGITS

 Towards

0.20

Towards

0.19

 toward

0.18

Tow

0.18

yb

0.16

Err

0.16

 towards

0.16

 prov

0.15

roperties

0.15

 robust

0.15

Activations Density 0.024%

unique numerical data and references related to scientific studies or methodologies

No Comments

No Known Activations