INDEX

Explanations

tokens that indicate significant actions or states of existence

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Solución

-0.42

Становништво

-0.39

 Weiterbildung

-0.36

 failures

-0.35

AndroidJUnit

-0.34

嘛

-0.34

 tilbage

-0.33

 ihan

-0.33

 failure

-0.33

 romanzo

-0.32

POSITIVE LOGITS

httphttps

0.57

PointerException

0.53

 autorytatywna

0.50

Rptr

0.50

 postIndex

0.49

 لينك

0.49

رشف

0.49

oneofs

0.48

 protoimpl

0.47

 Rptr

0.47

Activations Density 0.001%

tokens that indicate significant actions or states of existence

No Comments

No Known Activations