INDEX

Explanations

the letter "l"

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

tonode

-0.61

 telefónica

-0.60

 pescoço

-0.60

новниш

-0.50

UnitTesting

-0.50

 pouvoit

-0.50

 normalidad

-0.50

principalColumn

-0.49

 votación

-0.48

appartamento

-0.48

POSITIVE LOGITS

 commercial

0.51

hoeddwyd

0.50

tonsoft

0.48

 فريبيس

0.48

Autoritní

0.46

WriteTagHelper

0.45

commercial

0.45

 commer

0.44

 ErrIntOverflow

0.43

Bibliograf

0.41

Activations Density 0.000%

the letter "l"

No Comments

No Known Activations