INDEX

Explanations

terms and concepts related to academic and research-focused discussions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

aye

-0.15

ÑĢÐ°Ðº

-0.14

tah

-0.14

panic

-0.14

oui

-0.14

ilarity

-0.13

anoia

-0.13

prites

-0.13

Ã¡ct

-0.13

POSITIVE LOGITS

cie

0.15

 testim

0.15

allen

0.14

ÑĸÐºÐ°

0.14

 Hart

0.14

ãĥ¼ãĥĲ

0.14

ogan

0.14

erne

0.14

nes

0.13

leton

0.13

Activations Density 0.005%

terms and concepts related to academic and research-focused discussions

No Comments

No Known Activations