INDEX

Explanations

words indicating conditions or criteria related to requirements and implications

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

to determine or dictate

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

Empereur

-0.49

minecraftforge

-0.48

Ведь

-0.46

 Cæsar

-0.44

 PeEnEo

-0.44

mogat

-0.41

└──

-0.41

Assemblée

-0.41

bildēt

-0.41

 Confeder

-0.41

POSITIVE LOGITS

 nakalista

0.58

IVEREF

0.46

 esternos

0.45

tawesome

0.41

 насељу

0.39

RenderAtEndOf

0.39

 تضيفلها

0.38

 незавершена

0.38



0.37

IsContent

0.36

Activations Density 0.029%

words indicating conditions or criteria related to requirements and implications

to determine or dictate

No Comments

No Known Activations