INDEX

Explanations

phrases indicating assistance, support, or actions that lead to positive outcomes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 25-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_25/width_16k/average_l0_116

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.25.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

 magasiner

-0.73

 Initialized

-0.64

 catalyzed

-0.64

dized

-0.63

 honors

-0.61

 așa

-0.60

 discolor

-0.58

 honored

-0.58

 colors

-0.56

ized

-0.56

POSITIVE LOGITS

 mobili

1.34

 stabili

1.33

 sympathi

1.19

 standar

1.06

 specialise

1.02

 synthe

1.01

 saniti

1.01

 capital

1.01

 minimise

1.00

 prioritise

1.00

Activations Density 0.614%

phrases indicating assistance, support, or actions that lead to positive outcomes

No Comments

No Known Activations