INDEX

Explanations

concepts related to research and its implications

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

Follows "the", "field", "one", or "not"

np_acts-logits-general · gemini-2.0-flash

general statements, emphasis

np_acts-logits-general · gemini-2.5-flash-lite

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

styleType

-0.44

 Tata

-0.38

 salivary

-0.37

DoubleQuotes

-0.37

 anhyd

-0.36

 defStyle

-0.36

XtraReports

-0.36

onnaissance

-0.35

Packard

-0.35

 anch

-0.35

POSITIVE LOGITS

 particularmente

0.60

 للاسماء

0.54

 especialmente

0.52

 notamment

0.52

FieldOffset

0.50

jména

0.49

 particularly

0.49

 Dabei

0.49

 dabei

0.48

 notably

0.46

Activations Density 0.922%

concepts related to research and its implications

Follows "the", "field", "one", or "not"

general statements, emphasis

No Comments

No Known Activations