INDEX

Explanations

terms related to historical and scientific analysis, particularly in the context of research and implications in various fields

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

election analysis

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

Newswire

-0.61

ViewImports

-0.59

 propOrder

-0.57

GOTREF

-0.54

MENAFN

-0.54

 فريبيس

-0.54

ScopeManager

-0.53

MigrationBuilder

-0.52

 חיצוניים

-0.50

MessageOf

-0.50

POSITIVE LOGITS

 acuer

0.37

 AssemblyTitle

0.35

hyrchwyd

0.34

nehin

0.34

 láser

0.32

 asiático

0.31

 memang

0.30

OpenHelper

0.30

 paja

0.30

entlich

0.29

Activations Density 0.178%

terms related to historical and scientific analysis, particularly in the context of research and implications in various fields

election analysis

No Comments

No Known Activations