INDEX

Explanations

terms related to academic or formal descriptions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

leyin

-0.15

bote

-0.15

ä¸¸

-0.15

essa

-0.15

 Champ

-0.15

chas

-0.14

 reco

-0.14

ocaly

-0.14

 Bylo

-0.14

Integral

-0.14

POSITIVE LOGITS

rh

0.24

 males

0.24

lob

0.23

 port

0.23

 imper

0.23

pul

0.23

 conv

0.23

 grav

0.23

viv

0.23

 fauc

0.23

Activations Density 0.004%

terms related to academic or formal descriptions

No Comments

No Known Activations