INDEX

Explanations

expressions or discussions about controversial or inappropriate statements and their impacts

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

Words, remarks, or statements

np_acts-logits-general · gemini-2.0-flash

spoken words and statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 Polres

-0.54

photobucket

-0.53

 namanya

-0.53

navn

-0.51

Name

-0.49

 modello

-0.45

 bernama

-0.43

 indicato

-0.43

 Name

-0.43

ksessa

-0.42

POSITIVE LOGITS

 uttered

1.19

 heard

0.98

 reverber

0.90

 circulating

0.90

 overheard

0.84

 circulated

0.83

 echoing

0.82

 echoed

0.81

 spoken

0.81

SharedCtor

0.79

Activations Density 0.455%

expressions or discussions about controversial or inappropriate statements and their impacts

Words, remarks, or statements

spoken words and statements

No Comments

No Known Activations