INDEX

Explanations

specific terminology and keywords related to various contexts, such as legal, geographical, or descriptive terms

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

tokens followed by specific words

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

 fashiola

-0.79

bootstrapcdn

-0.76

AddTagHelper

-0.76

DockStyle

-0.72

SequentialGroup

-0.72

IUrlHelper

-0.71

MLLoader

-0.71

 defaultstate

-0.71

ſicht

-0.71

ConstraintMaker

-0.70

POSITIVE LOGITS

 Constitución

0.31

 assistência

0.27

 práctica

0.27

 recomendación

0.26

 protección

0.25

 Voraussetzung

0.25

 First

0.24

 spowod

0.24

 ESPECIAL

0.24

 segü

0.23

Activations Density 0.003%

specific terminology and keywords related to various contexts, such as legal, geographical, or descriptive terms

tokens followed by specific words

No Comments

No Known Activations