INDEX

Explanations

instances of the word "seen" in various contexts

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

seen followed by observation details

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

Not in Any Lists

No Comments

Negative Logits

 pleaſure

-0.67

 ſche

-0.54

ſelf

-0.53

 purpoſe

-0.53

 houſe

-0.50

 Majefty

-0.48

 myſelf

-0.47

izarse

-0.46

ellido

-0.46

fptr

-0.46

POSITIVE LOGITS

 seen

0.69

Seen

0.65

seen

0.65

见过

0.63

 Seen

0.57

 SEEN

0.57

看过

0.57

 been

0.57

haped

0.55

proven

0.54

Activations Density 0.008%

instances of the word "seen" in various contexts

seen followed by observation details

No Comments

No Known Activations