INDEX

Explanations

references to the name "Philip" and its variants

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

philip and related names

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

[]:

-0.57

 ***!

-0.56

WebRequest

-0.54

cray

-0.54

 Cray

-0.52

Gn

-0.52

/**

-0.52

corrhi

-0.50

}")

-0.50

CodeAttribute

-0.50

POSITIVE LOGITS

 Philipp

0.71

 Philip

0.67

 Phillip

0.66

 Philips

0.64

Philip

0.64

Phillip

0.63

Philipp

0.63

Phil

0.58

 Phil

0.57

 exécu

0.57

Activations Density 0.007%

references to the name "Philip" and its variants

philip and related names

No Comments

No Known Activations