INDEX

Explanations

references to threads, views, and related workspace content in a programming context

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

Tokens before names or IDs

np_acts-logits-general · gemini-2.0-flash

entity name or ID

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

IntoConstraints

-0.72

 resourceCulture

-0.66

WriteTagHelper

-0.60

extAlignment

-0.56

IBarButtonItem

-0.56

ínű

-0.53

IsContent

-0.53

 femmin

-0.52

obel

-0.52

цезда

-0.51

POSITIVE LOGITS

 name

0.84

name

0.78

Name

0.67

wide

0.63

 owner

0.57

ID

0.57

 info

0.55

ID

0.54

naam

0.53

 whose

0.53

Activations Density 0.595%

references to threads, views, and related workspace content in a programming context

Tokens before names or IDs

entity name or ID

No Comments

No Known Activations

references to threads, views, and related workspace content in a programming context

Tokens before names or IDs

entity name or ID

No Comments

No Known Activations