INDEX

Explanations

instances of the word "per" in various contexts, often related to measurements or rates

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

yard

-0.17

ä»¬

-0.15

ship

-0.15

elastic

-0.14

ek

-0.14

sel

-0.14

hani

-0.13

gili

-0.13

scape

-0.13

yr

-0.13

POSITIVE LOGITS

 annum

0.20

missive

0.19

malink

0.19

pendicular

0.18

iphery

0.17

ï¸ı

0.16

imeters

0.16

sea

0.16

cents

0.15

 capita

0.15

Activations Density 0.033%

instances of the word "per" in various contexts, often related to measurements or rates

No Comments

No Known Activations