INDEX

Explanations

terms related to reversing or negating actions or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_0/width_16k/average_l0_105

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

 vertellen

-0.56

 remplir

-0.54

 presentazione

-0.54

 detto

-0.52

 répé

-0.51

 treinamento

-0.51

Lähteet

-0.51

 remplacé

-0.51

 använder

-0.50

 puissiez

-0.50

POSITIVE LOGITS

 Processes

0.96

 processes

0.95

 Proc

0.92

Processes

0.86

Proc

0.85

processes

0.81

setVerticalGroup

0.78

 Process

0.76

 process

0.74

 PROCESSES

0.74

Activations Density 0.146%

terms related to reversing or negating actions or conditions

No Comments

No Known Activations