INDEX

Explanations

mentions of specific geographic locations or entities associated with geographic names

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

ARGIN

-0.18

zent

-0.17

zar

-0.16

udent

-0.15

hani

-0.15

rome

-0.15

plural

-0.15

role

-0.14

cessive

-0.14

POSITIVE LOGITS

ircon

0.24

ebra

0.23

onal

0.23

OOM

0.22

orro

0.22

abbix

0.21

entral

0.20

onedDateTime

0.20

estimate

0.19

Activations Density 0.032%

mentions of specific geographic locations or entities associated with geographic names

No Comments

No Known Activations