INDEX

Explanations

punctuation marks and various types of formatting

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

Clifford gave a natural generalization

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

principalColumn

-0.82

AddTagHelper

-0.79

ValueStyle

-0.75

хьтан

-0.72

 Italijani

-0.72

RTLR

-0.69

 useAppContext

-0.69

ロウィン

-0.67

 NSCoder

-0.65

 queſta

-0.65

POSITIVE LOGITS

落

0.38

Program

0.38

 Program

0.38

Running

0.36

 Schreiber

0.35

 Spector

0.35

 Zweifel

0.34

Alors

0.33

 Today

0.33

 This

0.32

Activations Density 0.005%

punctuation marks and various types of formatting

Clifford gave a natural generalization

No Comments

No Known Activations