INDEX

Explanations

titles and positions within governmental organizations, particularly related to historical political figures and structures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

Å¾el

-0.17

ea

-0.16

enses

-0.14

 Trem

-0.14

iam

-0.13

 pipe

-0.13

hir

-0.13

fat

-0.13

edis

-0.13

GridColumn

-0.13

POSITIVE LOGITS

rabbit

0.15

creature

0.15

deen

0.14

 åĬ

0.14

reak

0.14

mana

0.13

ynet

0.13

anks

0.13

 civilian

0.13

Activations Density 0.072%

titles and positions within governmental organizations, particularly related to historical political figures and structures

No Comments

No Known Activations