INDEX

Explanations

proper nouns, particularly names and locations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

Std

-0.15

à¸°à¹ģ

-0.15

ully

-0.15

 Ð¤ÐµÐ´ÐµÑĢÐ°ÑĨÐ¸Ð¸

-0.14

Sor

-0.14

 Banc

-0.14

 Bucc

-0.13

ius

-0.13

hti

-0.13

kaf

-0.13

POSITIVE LOGITS

 Straw

0.17

 æĸ

0.16

 itself

0.15

 samo

0.14

_REFRESH

0.14

 reap

0.14

eson

0.14

 ÑģÐ¾Ð±Ñĸ

0.14

ī

0.14

 Finder

0.14

Activations Density 0.255%

proper nouns, particularly names and locations

No Comments

No Known Activations