INDEX

Explanations

specific phrases or concepts related to lists, trails, data manipulation, and relationships

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

pandas, data science, and programming concepts

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

 Bohn

-0.45

 ligiloj

-0.41

 organization

-0.34

KURZBESCHREIBUNG

-0.34

 oprot

-0.34

Gym

-0.33

 Thompson

-0.33

 saumon

-0.33

 veter

-0.33

Eng

-0.32

POSITIVE LOGITS

 الحره

0.61

CodeAttribute

0.58

complexContent

0.57

InitVars

0.56

 فريبيس

0.54

GraphicsUnit

0.52

TypedDataSet

0.52

msgTypes

0.52

FieldDescriptor

0.51

 umana

0.51

Activations Density 0.400%

specific phrases or concepts related to lists, trails, data manipulation, and relationships

pandas, data science, and programming concepts

No Comments

No Known Activations