INDEX

Explanations

concepts related to theoretical frameworks and their underlying principles in various contexts

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

states a conclusion

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

GOTREF

-0.79

ConstraintMaker

-0.79

хьтан

-0.78

 queſta

-0.75

tvguidetime

-0.74

InputBorder

-0.72

webElementGuid

-0.72

 Wikimedijinoj

-0.71

<pad>

-0.70

<unused3>

-0.70

POSITIVE LOGITS

 this

0.38

 that

0.33

 sauter

0.31

この

0.29

 these

0.29

ot

0.28

 principes

0.28

Jay

0.28

tie

0.27

put

0.26

Activations Density 0.111%

concepts related to theoretical frameworks and their underlying principles in various contexts

states a conclusion

No Comments

No Known Activations