INDEX

Explanations

representations of class names in a programming context

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

code, files, and data

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

 tény

-0.38

 Administração

-0.38

 reír

-0.38

muñeca

-0.38

 vergüenza

-0.37

 považ

-0.36

 imaginación

-0.36

 disfrut

-0.36

 nepř

-0.36

 enju

-0.35

POSITIVE LOGITS

+#+

0.87

AutoScaleMode

0.73

ChildScrollView

0.72

TypedDataSet

0.71

 autorytatywna

0.69

WebServlet

0.66

iſche

0.60

expandindo

0.58

iſchen

0.58

 MaterialApp

0.57

Activations Density 0.021%

representations of class names in a programming context

code, files, and data

No Comments

No Known Activations