INDEX

Explanations

key terms indicating actions or expressions related to competitions or achievements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 7-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L7R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.7.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

 Mein

-0.15

upert

-0.15

.opendaylight

-0.14

iko

-0.14

/ge

-0.14

 sitesinde

-0.13

onne

-0.13

warts

-0.13

å§ĵ

-0.13

&id

-0.13

POSITIVE LOGITS

addock

0.15

azzi

0.15

à¸¹à¹Ī

0.15

ãģ¨ãĤĤ

0.14

ts

0.13

 Decomp

0.13

 ðŁ

0.13

é¥

0.13

BI

0.13

Activations Density 0.035%

key terms indicating actions or expressions related to competitions or achievements

No Comments

No Known Activations