INDEX

Explanations

instances of the word "on" in various contexts related to locations or directions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

-0.16

ettel

-0.16

abic

-0.15

rud

-0.14

ark

-0.14

ks

-0.14

 areas

-0.14

adesh

-0.13

 Ð¾Ð±Ð»Ð°ÑģÑĤ

-0.13

POSITIVE LOGITS

 stil

0.23

 Route

0.21

 wheels

0.19

 route

0.18

 opposite

0.17

aje

0.16

Route

0.15

Ø¢ÙħØ¯

0.15

 Wheels

0.15

ãĥ¬ãĤ¹

0.15

Activations Density 0.115%

instances of the word "on" in various contexts related to locations or directions

No Comments

No Known Activations