INDEX

Explanations

phrases and references to decision-making processes and factual clarifications in legal contexts

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 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

⤹

-0.46

 Савезне

-0.41

Disliked

-0.41

 ModelRenderer

-0.41

MethodType

-0.40

tagext

-0.38

oneofs

-0.38

writerow

-0.36

thâu

-0.36

 elab

-0.35

POSITIVE LOGITS

Hentet

0.60

 AspNetCore

0.57

ثيق

0.55

rungsseite

0.54

Хьажоргаш

0.53

AccessorTable

0.52

 يتيمه

0.51



0.51

 صوتيه

0.51

 jScrollPane

0.49

Activations Density 0.505%

phrases and references to decision-making processes and factual clarifications in legal contexts

No Comments

No Known Activations