INDEX

Explanations

punctuation marks, specifically commas and periods

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

otto

-0.16

aises

-0.15

 Ð³ÑĢÐ¸

-0.14

bff

-0.14

ieri

-0.14

aise

-0.14

amburger

-0.14

igner

-0.13

amat

-0.13

evi

-0.13

POSITIVE LOGITS

(~(

0.15

åģ

0.14

ANJI

0.14

 dispatch

0.14

untu

0.14

extensions

0.13

ATTERN

0.13

Ø§Ø¦ÙĤ

0.13

ongyang

0.13

Dispatch

0.13

Activations Density 0.181%

punctuation marks, specifically commas and periods

No Comments

No Known Activations