INDEX

Explanations

patterns related to numerical data or code structure

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

abel

-0.16

ä¿Ĭ

-0.16

urch

-0.15

amic

-0.15

ihn

-0.14

 Merry

-0.14

 gang

-0.14

_:*

-0.14

 Gang

-0.14

angler

-0.14

POSITIVE LOGITS

teki

0.15

tah

0.14

brtc

0.14

arov

0.14

 Gauge

0.13

-ln

0.13

bÃŃr

0.13

æĭ©

0.13

á»Ľp

0.13

walls

0.13

Activations Density 0.004%

patterns related to numerical data or code structure

No Comments

No Known Activations