INDEX

Explanations

terms related to measurement or analysis

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

istra

-0.19

anou

-0.16

achers

-0.15

alach

-0.15

oga

-0.15

Ã¼hl

-0.15

arez

-0.15

anja

-0.15

oice

-0.15

anj

-0.15

POSITIVE LOGITS

 Pitch

0.17

 micron

0.15

 Moran

0.15

Pitch

0.15

 Glass

0.15

 Meng

0.15

 Darling

0.15

 pitched

0.15

 fork

0.15

 fold

0.15

Activations Density 0.006%

terms related to measurement or analysis

No Comments

No Known Activations