INDEX

Explanations

references to specific people or concepts, particularly emphasizing the word "that."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 19-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_19/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.19.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

 geslacht

-0.55

 parachoque

-0.53

 posiblemente

-0.51

 gepubliceerd

-0.51

 pertanto

-0.49

 motifs

-0.49

 efectu

-0.48

 favoritas

-0.48

 regiments

-0.46

 tombs

-0.46

POSITIVE LOGITS

big

0.98

 huge

0.94

 wonderful

0.92

 pesky

0.91

 little

0.84

 fameux

0.82

 great

0.81

 extra

0.80

 large

0.80

 elusive

0.77

Activations Density 0.214%

references to specific people or concepts, particularly emphasizing the word "that."

No Comments

No Known Activations