INDEX

Explanations

references to political leaders and positions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 19-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L19R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.19.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

ennen

-0.18

ÄįnÃ½

-0.16

pij

-0.15

ajes

-0.14

ÅĻÃŃt

-0.14

vais

-0.14

reich

-0.14

eno

-0.14

 pardon

-0.14

amin

-0.13

POSITIVE LOGITS

 opposition

0.44

 Opposition

0.41

Opp

0.33

 opposing

0.27

 opposed

0.27

Opp

0.26

 oppose

0.25

 oppos

0.24

 Shadow

0.24

 Leader

0.24

Activations Density 0.051%

references to political leaders and positions

No Comments

No Known Activations