INDEX

Explanations

references to courtroom proceedings and legal terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 EconPapers

-0.88

 pleaſure

-0.75

windowFixed

-0.71

 fubject

-0.71

 Monfieur

-0.68

 iſt

-0.66

 hâte

-0.65

Rhestr

-0.65

tagHelperRunner

-0.63

URLException

-0.62

POSITIVE LOGITS

 told

0.99

 informed

0.88

 diberi

0.87

 receive

0.82

被

0.77

told

0.75

 receives

0.73

 worden

0.73

 Told

0.73

 receiving

0.71

Activations Density 0.244%

references to courtroom proceedings and legal terminology

No Comments

No Known Activations