INDEX

Explanations

references to high-ranking officials or titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

ctype

-0.16

Ã¶g

-0.15

atik

-0.15

ableView

-0.15

ubits

-0.15

áº¥u

-0.14

Äįas

-0.14

 Celt

-0.14

ynch

-0.14

eration

-0.14

POSITIVE LOGITS

dom

0.25

 executive

0.25

lain

0.22

 operating

0.22

 justice

0.21

 Whip

0.21

 executives

0.21

-exec

0.21

ëª

0.21

-of

0.20

Activations Density 0.014%

references to high-ranking officials or titles

No Comments

No Known Activations