INDEX

Explanations

specific locations or components associated with various objects and features

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

addy

-0.16

ca

-0.14

aeda

-0.14

shaw

-0.14

enen

-0.14

 statuses

-0.13

åķ

-0.13

.tie

-0.13

 planet

-0.13

POSITIVE LOGITS

jabi

0.17

tember

0.16

byname

0.16

atem

0.15

 damer

0.14

Å¡ti

0.14

ovky

0.14

ioni

0.14

Î¿ÏħÏĥ

0.14

obierno

0.14

Activations Density 0.285%

specific locations or components associated with various objects and features

No Comments

No Known Activations