INDEX

Explanations

programming-related terminology and structure definitions

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

programming language constructs

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

 CreateTagHelper

-0.70

RegressionTest

-0.68

 المعيارى

-0.66

DoubleQuotes

-0.65

principalTable

-0.64

HostException

-0.61

GraphicsUnit

-0.60

 محفوظة

-0.58

 Италијани

-0.58

jména

-0.57

POSITIVE LOGITS

 CanadaChoose

0.34

 untuk

0.31

ája

0.30

for

0.30

 связан

0.29

verture

0.28

 fallait

0.28

 logic

0.28

untuk

0.28

 voor

0.28

Activations Density 0.002%

programming-related terminology and structure definitions

programming language constructs

No Comments

No Known Activations