INDEX

Explanations

references to printing and code execution within a programming context

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

"directly" or its foreign translations

np_acts-logits-general · gemini-2.0-flash

directly integrated functionality

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_16k/average_l0_76

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.31.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

apimachinery

-0.47

withIdentifier

-0.40

InitStructure

-0.37

gnet

-0.36

 resourceCulture

-0.36

 beginnetje

-0.35

"<?

-0.35

complexContent

-0.35

ępo

-0.35

ensee

-0.35

POSITIVE LOGITS

 directly

2.14

directly

2.05

 directement

1.97

 directamente

1.91

 Directly

1.88

 direttamente

1.88

直接

1.77

 diretamente

1.77

 bezpośrednio

1.70

 langsung

1.66

Activations Density 0.600%

references to printing and code execution within a programming context

"directly" or its foreign translations

directly integrated functionality

No Comments

No Known Activations