INDEX

Explanations

statements that express subjective assessments or recommendations regarding actions and decisions

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

Tokens preceding "leave"

np_acts-logits-general · gemini-2.0-flash

leave or finish

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

AndEndTag

-0.59

elcome

-0.52

XmlAccessType

-0.51

Mismatch

-0.51

UnknownFieldSet

-0.51

ragan

-0.48

icide

-0.48

Personensuche

-0.48

outSlope

-0.48

 Vitale

-0.46

POSITIVE LOGITS

 leave

2.45

 Leave

2.34

Leave

2.25

leave

2.19

 leaving

2.13

 LEAVE

1.94

leaving

1.90

 Leaving

1.81

 left

1.80

LEAVE

1.79

Activations Density 0.416%

statements that express subjective assessments or recommendations regarding actions and decisions

Tokens preceding "leave"

leave or finish

No Comments

No Known Activations