INDEX

Explanations

references to political processes and elections

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

DST

-0.17

olicy

-0.16

verb

-0.15

DST

-0.14

uum

-0.14

trap

-0.14

levation

-0.14

din

-0.14

pants

-0.14

policy

-0.13

POSITIVE LOGITS

 Voters

0.28

 voters

0.28

 polling

0.28

 voter

0.27

 Poll

0.25

 Voter

0.25

 poll

0.24

 voting

0.23

 electoral

0.23

 ballot

0.22

Activations Density 0.014%

references to political processes and elections

No Comments

No Known Activations