INDEX

Explanations

references to sand or sandy environments

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

sand and related concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

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

Negative Logits

hape

-0.62

songwriter

-0.59

tabular

-0.58

 Alonzo

-0.57

 Hadley

-0.57

 déri

-0.56

lihood

-0.55

mės

-0.52

Herald

-0.52

Cyfeiriadau

-0.51

POSITIVE LOGITS

 dunes

0.96

wic

0.91

 dune

0.84

 beaches

0.78

 SAND

0.76

 Sand

0.73

Sand

0.73

SAND

0.71

 sand

0.69

Sanders

0.69

Activations Density 0.021%

references to sand or sandy environments

sand and related concepts

No Known Activations