INDEX

Explanations

elements related to sequences, including transformations and comparisons

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 35-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_35/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.35.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

 &___

-0.51

bentukan

-0.40

DotNetBar

-0.39

GOTREF

-0.38

těte

-0.37

 Wikimedijinoj

-0.35

 mourut

-0.35

 насељу

-0.34

IBOutlet

-0.33

 olímp

-0.33

POSITIVE LOGITS

 reverse

3.38

 reversed

2.97

 Reverse

2.88

reverse

2.78

 inverse

2.75

Reverse

2.69

 reverses

2.59

 reversing

2.50

逆

2.41

 reversal

2.38

Activations Density 1.776%

elements related to sequences, including transformations and comparisons

No Comments

No Known Activations