INDEX

Explanations

phrases that indicate past actions or experiences involving the word "have"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_21/width_16k/average_l0_70

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

now

-0.65

 often

-0.61

ViewInit

-0.61

 spesso

-0.58

 tantôt

-0.58

 ormai

-0.56

InstrumentedTest

-0.56

ParallelGroup

-0.54

ImageContext

-0.54

どうやら

-0.53

POSITIVE LOGITS

 hubiera

0.74

GEBURTSDATUM

0.72

Décès

0.69

expected

0.69

 hätte

0.68

Viited

0.68

 volna

0.68

 hubieran

0.67

—

0.67

 expected

0.66

Activations Density 0.127%

phrases that indicate past actions or experiences involving the word "have"

No Comments

No Known Activations