INDEX

Explanations

pronouns and punctuation indicating dialogue or quotations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

soever

-0.15

olib

-0.15

rok

-0.14

âĢĮÛĮ

-0.14

 scissors

-0.13

ëĭĿ

-0.13

_gradients

-0.13

eneric

-0.13

 Frid

-0.13

Â·

-0.13

POSITIVE LOGITS

zman

0.17

å·±

0.14

tas

0.14

insky

0.14

ureau

0.14

nah

0.14

acement

0.13

arda

0.13

jen

0.13

 ÄĳÃ³

0.13

Activations Density 0.002%

pronouns and punctuation indicating dialogue or quotations

No Comments

No Known Activations