INDEX

Explanations

repeated instances of specific letters in various contexts, particularly 'j' and 'o'

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

iyan

-0.17

agogue

-0.15

ODB

-0.15

iÃª

-0.15

iáº¿n

-0.15

ToJson

-0.15

jekt

-0.15

iÃ³

-0.15

uchos

-0.15

Ø§ÙħØ¨Ø±

-0.14

POSITIVE LOGITS

ournals

0.28

oints

0.23

ourn

0.23

à¥įà¤ŀ

0.22

ee

0.20

upyter

0.20

à¤¼

0.19

argon

0.19

ection

0.19

umper

0.18

Activations Density 0.107%

repeated instances of specific letters in various contexts, particularly 'j' and 'o'

No Comments

No Known Activations