INDEX

Explanations

quantitative descriptors relating to estimates or approximations of values

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

localctx

-0.60

emlrt

-0.57

 pinulongan

-0.57

ロウィン

-0.56

 MemoryStream

-0.55

majánló

-0.55

nub

-0.53

 PyTuple

-0.51

angdong

-0.51

lgari

-0.51

POSITIVE LOGITS

ConstraintMaker

0.44

Décès

0.38

 Sénat

0.36

IndentedString

0.35

 souverain

0.34

beyond

0.33

 coté

0.33

 beyond

0.32

 géant

0.31

 Administração

0.31

Activations Density 0.043%

quantitative descriptors relating to estimates or approximations of values

No Comments

No Known Activations