INDEX

Explanations

technical or programming-related terminology

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

ÏħÎ³

-0.19

ãĥĸãĥ«

-0.15

tile

-0.15

ÏħÏĩ

-0.15

 Tick

-0.14

 Tiles

-0.14

 Firearms

-0.14

 Tide

-0.14

 Tactical

-0.14

 Toolbox

-0.14

POSITIVE LOGITS

 tree

0.94

 trees

0.83

 Tree

0.81

tree

0.79

-tree

0.78

Tree

0.75

_tree

0.74

æłĳ

0.70

 Trees

0.70

.tree

0.69

Activations Density 0.225%

technical or programming-related terminology

No Comments

No Known Activations