INDEX

Explanations

specific nouns or proper nouns that indicate key elements or subjects in a discussion

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

orian

-0.17

Roe

-0.16

lez

-0.16

sel

-0.14

Hub

-0.14

Ses

-0.14

ãĤ¹ãĤ«

-0.14

Vo

-0.14

uden

-0.14

ium

-0.14

POSITIVE LOGITS

ãĥ»ãĥ»ãĥ»↵↵

0.17

atron

0.16

0.15

acket

0.15

AML

0.15

uin

0.15

otch

0.15

 sentinel

0.14

/lic

0.14

pent

0.14

Activations Density 0.020%

specific nouns or proper nouns that indicate key elements or subjects in a discussion

No Comments

No Known Activations