INDEX

Explanations

terms related to concepts of reconstruction, intervention, and legal proceedings

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

Begins with "con" or contains "recon"

np_acts-logits-general · gemini-2.0-flash

con- prefixes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 24-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_24/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.24.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

 Chrif

-0.63

helves

-0.60

databind

-0.58

 authorize

-0.57

 Ditto

-0.57

tize

-0.55

 ſta

-0.55

FORME

-0.55

 stepfather

-0.55

ESTONE

-0.54

POSITIVE LOGITS

version

0.55

 betweenstory

0.51

HasColumnType

0.49

Formazione

0.47

Cep

0.45

cep

0.44

fer

0.43

du

0.43

cur

0.42

ecu

0.42

Activations Density 0.450%

terms related to concepts of reconstruction, intervention, and legal proceedings

Begins with "con" or contains "recon"

con- prefixes

No Comments

No Known Activations