INDEX

Explanations

programming-related constructs, particularly function definitions and object instantiation

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

class constructor syntax

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

 threatening

-0.48

 frightening

-0.43

 distracting

-0.43

 disturbing

-0.42

 devastating

-0.41

kulu

-0.41

 pokok

-0.41

 gece

-0.40

 pictures

-0.40

 main

-0.40

POSITIVE LOGITS

تقاوى

0.83

 kasarigan

0.81

 tartalomajánló

0.71

Personendaten

0.67

IntoConstraints

0.61

 nakalista

0.60

 صوتيه

0.60

 invokingState

0.59

 awakeFromNib

0.57

rungsseite

0.56

Activations Density 0.001%

programming-related constructs, particularly function definitions and object instantiation

class constructor syntax

No Comments

No Known Activations