INDEX

Explanations

terms related to academic and theoretical concepts across various fields

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

 Tang

-0.14

å±Ģ

-0.14

rote

-0.14

 Minister

-0.13

onte

-0.13

 minister

-0.13

 Forced

-0.13

rol

-0.13

POSITIVE LOGITS

plier

0.15

hall

0.15

ansa

0.14

mium

0.14

chen

0.14

gulp

0.14

èĪį

0.14

ittest

0.14

stag

0.14

NewLabel

0.14

Activations Density 0.432%

terms related to academic and theoretical concepts across various fields

No Comments

No Known Activations