INDEX

Explanations

punctuation and formatting elements often found in metadata or time signatures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

agan

-0.15

ckt

-0.15

ç«

-0.15

ernaut

-0.15

ibri

-0.14

olumn

-0.14

URN

-0.13

rzy

-0.13

dod

-0.13

manager

-0.13

POSITIVE LOGITS

Î´Î®

0.16

FER

0.16

|required

0.14

IFORM

0.14

-runtime

0.14

éĹ

0.14

 track

0.13

âĢŀM

0.13

ISOString

0.13

ider

0.13

Activations Density 0.013%

punctuation and formatting elements often found in metadata or time signatures

No Comments

No Known Activations