INDEX

Explanations

references to significant places, ideas, or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

pedia

-0.15

vore

-0.15

lef

-0.15

ĩ¼

-0.15

vik

-0.15

æĸ¹

-0.14

WR

-0.14

jiÅ¡Å¥

-0.14

èŀº

-0.14

ØŃÙĦ

-0.14

POSITIVE LOGITS

orst

0.17

 BigNumber

0.15

aska

0.15

Å¡ek

0.14

ureka

0.14

 Klein

0.14

 file

0.14

mlin

0.14

å¿µ

0.14

ove

0.14

Activations Density 0.014%

references to significant places, ideas, or entities

No Comments

No Known Activations