INDEX

Explanations

structured data representations, particularly related to time and numerical sequences

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

raki

-0.16

 ä¸ĥ

-0.15

theid

-0.15

 Fifth

-0.14

rael

-0.14

rack

-0.14

kes

-0.14

ader

-0.14

ÄħÅ¼

-0.14

å®ħ

-0.14

POSITIVE LOGITS

0.75

 nine

0.62

 ninth

0.59

ä¹Ŀ

0.59

 Nine

0.59

 ä¹Ŀ

0.57

 Ninth

0.57

0.56

nine

0.56

ï¼Ļ

0.52

Activations Density 0.076%

structured data representations, particularly related to time and numerical sequences

No Comments

No Known Activations