INDEX

Explanations

numeric values and statistical terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

appa

-0.17

 Compound

-0.15

ç

-0.15

 Dawn

-0.15

iece

-0.15

 Jacob

-0.15

 compound

-0.14

utsch

-0.14

Ø§Ø¦Ø±

-0.14

Compound

-0.14

POSITIVE LOGITS

 digit

0.69

 digits

0.63

 Digit

0.59

digit

0.59

Digit

0.57

-digit

0.54

digits

0.52

_digit

0.50

dig

0.49

DIG

0.49

Activations Density 0.056%

numeric values and statistical terms

No Comments

No Known Activations