INDEX

Explanations

references to organizations or groups, particularly government or institutional agencies

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

ØªØ±ÛĮ

-0.15

phant

-0.15

bject

-0.15

angen

-0.15

cie

-0.14

coming

-0.14

lej

-0.14

.infinity

-0.14

engin

-0.14

leo

-0.14

POSITIVE LOGITS

.uk

0.15

AAF

0.15

ekler

0.15

ixels

0.14

ical

0.14

altar

0.14

oyal

0.14

nbsp

0.14

à¸±à¸Ĺ

0.14

hood

0.14

Activations Density 0.034%

references to organizations or groups, particularly government or institutional agencies

No Comments

No Known Activations