INDEX

Explanations

objects and their connections or attachments in descriptions

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

about attaching or hanging something

np_acts-logits-general · gemini-2.0-flash

hanging and attachment points

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

AndEndTag

-0.49

iseries

-0.47

Slice

-0.47

енча

-0.43

PageIndex

-0.43

ssf

-0.42

cov

-0.41

uga

-0.41

ssen

-0.39

lito

-0.39

POSITIVE LOGITS

 hanging

0.99

 attachment

0.97

 hang

0.93

 treo

0.92

 hangs

0.91

 крепления

0.91

 hangers

0.91

 hanger

0.90

 hooks

0.90

 attach

0.89

Activations Density 0.272%

objects and their connections or attachments in descriptions

about attaching or hanging something

hanging and attachment points

No Comments

No Known Activations