INDEX

Explanations

references to a specific watchmaker's name

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

Ø§ÙħÛĮÙĨ

-0.18

 Holt

-0.14

urn

-0.14

ctl

-0.14

URN

-0.14

ä¸

-0.14

aldi

-0.14

lear

-0.14

 probs

-0.14

 Pearce

-0.13

POSITIVE LOGITS

.mods

0.15

inton

0.14

 cust

0.14

dac

0.13

ÙħÙĨ

0.13

asion

0.13

lemn

0.13

meli

0.13

cust

0.13

.Focused

0.13

Activations Density 0.081%

references to a specific watchmaker's name

No Comments

No Known Activations