INDEX

Explanations

conjunctions and punctuation that indicate choices or alternatives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_4/width_65k/average_l0_66

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.4.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

:✨

-0.61

Enllaces

-0.57

 pulso

-0.55

 '*':

-0.55

ResumeLayout

-0.54

ètres

-0.53

HasAnnotation

-0.52

 yoksa

-0.52

URG

-0.51

PreferredItem

-0.50

POSITIVE LOGITS

tagHelperRunner

0.57

 numerus

0.56

AxisAlignment

0.54

 فريبيس

0.54

↵

0.54

ANTLR

0.52

adpleegd

0.52

AutoresizingMask

0.52

Geplaatst

0.50

 ComVisible

0.50

Activations Density 0.058%

conjunctions and punctuation that indicate choices or alternatives

No Comments

No Known Activations