INDEX

Explanations

instances of the word "that" and its variations in various contexts

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

introduces descriptive clauses

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

 Aérea

-0.49

ambién

-0.48

ußt

-0.47

 Infór

-0.46

 Româ

-0.46

 Públicas

-0.43

 démocr

-0.43

inginan

-0.42

 fédé

-0.42

 stället

-0.42

POSITIVE LOGITS

stateMutability

0.58

 CreateTagHelper

0.57

 ویکی‌پدی

0.56

Климат

0.55

iNdEx

0.53

 محفوظة

0.53

BagConstraints

0.50

 mencakup

0.50

GIVEREF

0.50

AccessorTable

0.49

Activations Density 0.059%

instances of the word "that" and its variations in various contexts

introduces descriptive clauses

No Comments

No Known Activations