INDEX

Explanations

references to graduates and graduation

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

graduate of

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

 pins

-0.42

 Thing

-0.40

 Things

-0.40

 bits

-0.40

 boxes

-0.40

 Espagne

-0.40

Box

-0.39

tinyos

-0.39

wezig

-0.39

 Bits

-0.39

POSITIVE LOGITS

 graduate

1.09

 GRADUATE

1.03

 Graduate

1.03

 Graduates

1.00

Graduate

0.98

ADUATE

0.98

graduate

0.97

 graduates

0.88

 graduation

0.87

Graduation

0.83

Activations Density 0.004%

references to graduates and graduation

graduate of

No Comments

No Known Activations