INDEX

Explanations

references to coding or programming-related components and functions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 9-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_9/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.9.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

BASELINE

-0.75

 InputDecoration

-0.72

rungsseite

-0.65

 unknownFields

-0.64

 GenerationType

-0.61

SpringBootTest

-0.61

 cref

-0.60

IntoConstraints

-0.59

ectoria

-0.58

})();

-0.57

POSITIVE LOGITS

ſelves

0.79

 Diſ

0.75

 perſon

0.73

 ſmall

0.73

 ſtate

0.72

 ſever

0.72

 pleaſure

0.72

ſelf

0.71

 Reſ

0.68

 leſs

0.68

Activations Density 0.953%

references to coding or programming-related components and functions

No Comments

No Known Activations