INDEX

Explanations

terms related to official sources or entities

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

"official" preceding nouns

np_acts-logits-general · gemini-2.0-flash

official website

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

 שוליים

-0.79

 CreateTagHelper

-0.70

literally

-0.69

enumi

-0.65

expandindo

-0.62

 surla

-0.60

enumii

-0.58

 célèbres

-0.57

 literally

-0.57

PYX

-0.56

POSITIVE LOGITS

dom

0.80

OFFICIAL

0.76

ization

0.74

isation

0.69

 OFFICIAL

0.69

dehy

0.66

ising

0.66

ize

0.65

ized

0.65

적으로

0.64

Activations Density 0.096%

terms related to official sources or entities

"official" preceding nouns

official website

No Comments

No Known Activations