INDEX

Explanations

phrases emphasizing possession or existence in relation to people, places, or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

asl

-0.19

odash

-0.14

 Fant

-0.14

anno

-0.13

icia

-0.13

ius

-0.13

rouw

-0.13

 began

-0.13

ãĥ³ãĥĢ

-0.13

POSITIVE LOGITS

 long

0.21

lain

0.20

 weather

0.19

 stood

0.19

 served

0.19

 lain

0.19

 hosted

0.18

 existed

0.17

 held

0.17

 fascinated

0.17

Activations Density 0.085%

phrases emphasizing possession or existence in relation to people, places, or events

No Comments

No Known Activations