INDEX

Explanations

classes and modules in programming contexts

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

zw

-0.14

 Levin

-0.14

illy

-0.14

 Smile

-0.14

avl

-0.14

ogan

-0.14

vn

-0.14

onom

-0.13

appa

-0.13

edar

-0.13

POSITIVE LOGITS

SSERT

0.17

Ã©rt

0.15

BindingUtil

0.14

SetName

0.14

 Bark

0.14

ekil

0.14

ibble

0.13

Beer

0.13

undef

0.13

 Milk

0.13

Activations Density 0.004%

classes and modules in programming contexts

No Comments

No Known Activations