INDEX

Explanations

the UTF-8 character encoding in content type specifications and related contexts

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-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_0/width_65k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.0.hook_mlp_out

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

Efq

-1.13

 Reſ

-1.09

 Houſe

-1.08

 Diſ

-1.04

^(@)

-1.02

 Anſ

-1.02

 itſelf

-1.02

 Monfieur

-1.01

 Мексичка

-1.01

 raiſ

-1.00

POSITIVE LOGITS

("

0.90

('

0.84

0.82

de

0.75

↵

0.74

="

0.74

0.73

::

0.72

Activations Density 0.099%

the UTF-8 character encoding in content type specifications and related contexts

No Comments

No Known Activations