INDEX

Explanations

mentions of discussions, proposals, or theories related to policies or ideas

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

umas

-0.15

 Demand

-0.15

nia

-0.15

gne

-0.14

Ã¤hlen

-0.14

odesk

-0.14

åħ¥ãĤĬ

-0.14

º

-0.14

wargs

-0.13

lice

-0.13

POSITIVE LOGITS

 raised

0.37

 advanced

0.36

raised

0.30

 raise

0.29

advanced

0.29

 Raised

0.29

 Advanced

0.26

 rais

0.26

-ra

0.25

 raising

0.25

Activations Density 0.270%

mentions of discussions, proposals, or theories related to policies or ideas

No Comments

No Known Activations