INDEX

Explanations

instances of decision-making and actions taken in a context of authority or control

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

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

 >=",

-0.78

AccessorTable

-0.69

tamen

-0.66

GraphicsUnit

-0.65

MessageTagHelper

-0.64

Personensuche

-0.63

 EconPapers

-0.58

 otomatig

-0.57

tagHelperRunner

-0.56

 bezeichneter

-0.55

POSITIVE LOGITS

 directly

3.38

 direct

3.10

directly

3.01

 Directly

2.93

Direct

2.78

 directement

2.74

direct

2.73

 directamente

2.67

 direttamente

2.55

 diretamente

2.54

Activations Density 0.853%

instances of decision-making and actions taken in a context of authority or control

No Comments

No Known Activations