INDEX

Explanations

references to historical structures and their characteristics

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

 condol

-0.14

oro

-0.14

 condolences

-0.14

 rekl

-0.13

 Branch

-0.13

afd

-0.13

Affected

-0.13

æĥĳ

-0.13

META

-0.13

eger

-0.13

POSITIVE LOGITS

 fort

0.43

 defensive

0.34

 forts

0.33

fort

0.32

 Fort

0.32

 ramp

0.32

 defenses

0.31

 castle

0.30

Fort

0.30

 bast

0.29

Activations Density 0.171%

references to historical structures and their characteristics

No Comments

No Known Activations