INDEX

Explanations

closing brackets

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

closing brace or item

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

 betweenstory

-0.77

tvguidetime

-0.73

 queſta

-0.71

 disambiguazione

-0.70

ſſung

-0.68

AnimationsModule

-0.68

bootstrapcdn

-0.67

 beſch

-0.66

PyExc

-0.66

 geſch

-0.65

POSITIVE LOGITS

0.30

↵

0.28

 veut

0.26

↵↵↵

0.26

</b>

0.25

 veulent

0.25

 tarda

0.25

↵↵

0.24

<h1>

0.24

eeu

0.24

Activations Density 0.001%

closing brackets

closing brace or item

No Comments

No Known Activations