INDEX

Explanations

code snippets and terminology, with some sensitivity to descriptions of landscapes

oai_token-act-pair · gemini-2.0-flash

code/hanging

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_4/width_16k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.ln2.hook_normalized

Architecture

jumprelu_transcoder

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 walls

-1.01

 wall

-1.00

 fences

-0.98

 fence

-0.97

 barriers

-0.87

Walls

-0.85

 Fence

-0.84

 WALL

-0.84

Wall

-0.84

walls

-0.84

POSITIVE LOGITS

 gambe

0.57

CompilerServices

0.51

 spared

0.50

 jornadas

0.50

 dess

0.49

φη

0.48

 Edson

0.48

rungsseite

0.47

 choisissez

0.47

 vientre

0.47

Activations Density 1.694%

code snippets and terminology, with some sensitivity to descriptions of landscapes

code/hanging

No Comments

No Known Activations