INDEX

Explanations

percentage-related terms and formatting symbols

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

C code snippets with error messages

np_acts-logits-general · gemini-2.0-flash

format specifiers in logging

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

ViewFeatures

-0.62

IUrlHelper

-0.60

WritableDatabase

-0.59

Personendaten

-0.55

RenderAtEndOf

-0.55

ScopeManager

-0.55

"}")

-0.53

NavController

-0.52

[];

-0.51

Rhestr

-0.51

POSITIVE LOGITS

SBATCH

0.73

AxisAlignment

0.57

UnknownFieldSet

0.55

 Ambiental

0.54

herited

0.52

脚注の使い方

0.52

GeoNames

0.52

 binoculars

0.52

ธอ

0.51

 Brotherhood

0.51

Activations Density 0.042%

percentage-related terms and formatting symbols

C code snippets with error messages

format specifiers in logging

No Comments

No Known Activations