INDEX

Explanations

references to political party affiliations, particularly focusing on the Republican Party

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 27-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L27R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.27.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

 bipartisan

-0.19

 Democrat

-0.17

 Democrats

-0.16

 democracy

-0.16

 Democratic

-0.16

 Liberals

-0.15

 diplomacy

-0.15

 republican

-0.15

 democratic

-0.14

POSITIVE LOGITS

 Party

0.37

-controlled

0.29

Party

0.28

-leaning

0.25

 controlled

0.23

-held

0.23

-led

0.21

 PARTY

0.21

-dominated

0.20

controlled

0.20

Activations Density 0.029%

references to political party affiliations, particularly focusing on the Republican Party

No Comments

No Known Activations