INDEX

Explanations

occurrences of terms related to unique identifiers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.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

rim

-0.15

fax

-0.15

ined

-0.15

ëģĶ

-0.15

aken

-0.15

elen

-0.15

mar

-0.15

yer

-0.14

roe

-0.14

Cec

-0.14

POSITIVE LOGITS

deen

0.17

 ë´ī

0.16

ensi

0.15

*>(&

0.15

 Andersen

0.15

0.14

ohana

0.14

cene

0.14

wine

0.14

Activations Density 0.011%

occurrences of terms related to unique identifiers

No Comments

No Known Activations