INDEX

Explanations

references to cited sources or evidence in a text

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

citing or cited examples

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

MigrationBuilder

-0.45

 houſe

-0.38

you

-0.38

ollection

-0.36

 civi

-0.35

PerformLayout

-0.35

Reparto

-0.34

QApplication

-0.34

 corpo

-0.32

 Criminal

-0.32

POSITIVE LOGITS

 examples

0.72

SequentialGroup

0.71

 Beispiele

0.69

 exemplos

0.68

 Signalez

0.65

 esimer

0.63

 citing

0.63

 cited

0.63

 CITED

0.62

 esempi

0.62

Activations Density 0.025%

references to cited sources or evidence in a text

citing or cited examples

No Comments

No Known Activations