INDEX

Explanations

terms related to time or sequence, particularly those indicating positions like "before" or "after."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 16-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L16R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.16.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

ephy

-0.16

Ð¼ÐµÑĩ

-0.15

arella

-0.15

abwe

-0.15

agu

-0.15

fid

-0.15

thro

-0.14

 Animalia

-0.14

vro

-0.14

zing

-0.14

POSITIVE LOGITS

most

0.15

acha

0.15

weeney

0.15

 Constantin

0.14

 definitive

0.14

beh

0.14

tors

0.14

echa

0.14

arl

0.14

noon

0.14

Activations Density 0.058%

terms related to time or sequence, particularly those indicating positions like "before" or "after."

No Comments

No Known Activations