INDEX

Explanations

programming language syntax and function-related terms

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

input operations in code

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

expandindo

-0.60

 BoxFit

-0.60

 &___

-0.58

RegressionTest

-0.53

 utafitiHapana

-0.49

AddHtmlAttribute

-0.48

 الحره

-0.47

 Photocase

-0.47

GEBURTS

-0.47

 виправивши

-0.47

POSITIVE LOGITS

stdc

0.56

cstdio

0.49

ImageField

0.43

iale

0.36

AMO

0.34

iriman

0.33

küche

0.32

AutoSize

0.32

bufio

0.32

DataInputStream

0.32

Activations Density 0.002%

programming language syntax and function-related terms

input operations in code

No Comments

No Known Activations