INDEX

Explanations

code-related terms and functions, particularly those associated with display management and extensions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

.MoveNext

-0.15

unik

-0.15

á»ķi

-0.14

ÑĢÑĥÐ½

-0.14

 Monte

-0.14

 workforce

-0.14

Eigen

-0.14

 passages

-0.14

SystemService

-0.14

rian

-0.13

POSITIVE LOGITS

 Atom

0.23

Atom

0.23

 atoms

0.23

 atom

0.23

atoms

0.22

 colormap

0.21

atom

0.21

Atoms

0.21

ATOM

0.21

_atoms

0.20

Activations Density 0.065%

code-related terms and functions, particularly those associated with display management and extensions

No Comments

No Known Activations