INDEX

Explanations

components and syntax related to programming or markup languages

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

Tokens often followed by "label" or "Label"

np_acts-logits-general · gemini-2.0-flash

edge label is plaintiff

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 متعلقه

-0.84

Portail

-0.65

FormTagHelper

-0.60

 mergeFrom

-0.60

QtCore

-0.57

पया

-0.55

 QtCore

-0.55

isContained

-0.54

RegressionTest

-0.52

iprot

-0.51

POSITIVE LOGITS

 label

3.04

label

2.73

 Label

2.73

 labels

2.71

Label

2.62

 LABEL

2.59

 labeling

2.46

 Labels

2.42

LABEL

2.42

 labelling

2.31

Activations Density 0.079%

components and syntax related to programming or markup languages

Tokens often followed by "label" or "Label"

edge label is plaintiff

No Comments

No Known Activations