INDEX

Explanations

phrases related to fitting or alignment within contexts or themes

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

 desmotivaciones

-0.59

 judiciaire

-0.58

gegangen

-0.54

 procès

-0.53

 indígen

-0.53

 judicia

-0.52

 fidé

-0.52

 وتسجيلات

-0.51

 bezpečnost

-0.49

 künst

-0.48

POSITIVE LOGITS

 perfect

0.57

 suited

0.56

 perfectly

0.54

suited

0.54

perfect

0.51

 PERFECT

0.50

 suitable

0.48

Perfect

0.47

 Perfect

0.46

 phù

0.46

Activations Density 0.444%

phrases related to fitting or alignment within contexts or themes

No Comments

No Known Activations