INDEX

Explanations

references to political candidates or electoral activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

.rmi

-0.16

 Vote

-0.16

andin

-0.16

vard

-0.15

dle

-0.15

kud

-0.15

Vote

-0.15

cul

-0.15

croft

-0.15

vote

-0.14

POSITIVE LOGITS

clid

0.16

0.15

oot

0.15

iming

0.15

inos

0.15

his

0.15

Haz

0.14

Themes

0.14

ghost

0.14

Haz

0.14

Activations Density 0.045%

references to political candidates or electoral activities

No Comments

No Known Activations