INDEX

Explanations

text that includes code and programming-related instructions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

nt

-0.15

ute

-0.14

't

-0.14

"+"

-0.13

ucks

-0.13

()

-0.13

ling

-0.13

quential

-0.13

lie

-0.13

ero

-0.13

POSITIVE LOGITS

 endregion

0.20

 TODO

0.15

TODO

0.15

ÙģØ³

0.14

filer

0.14

ikal

0.14

":[{↵

0.14

 FIXME

0.14

 =============================================================================↵

0.13

Activations Density 0.086%

text that includes code and programming-related instructions

No Comments

No Known Activations