INDEX

Explanations

the word "partly" and its variations, indicating an emphasis on partial explanations or causes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

cha

-0.19

ics

-0.17

chip

-0.15

ul

-0.14

iat

-0.14

yo

-0.14

pack

-0.14

zcze

-0.14

POSITIVE LOGITS

/full

0.17

äºľ

0.16

Ama

0.16

 createContext

0.14

atoon

0.14

_decorator

0.14

bilt

0.14

 /*!<

0.14

OOK

0.13

isans

0.13

Activations Density 0.010%

the word "partly" and its variations, indicating an emphasis on partial explanations or causes

No Comments

No Known Activations