INDEX

Explanations

key information about company or organization names and their related activities

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

followed by "was"

np_acts-logits-general · gemini-2.0-flash

official announcements and filings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 24-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_24/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.24.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

 siempre

-0.87

 hasn

-0.84

 always

-0.82

always

-0.77

siempre

-0.76

 restent

-0.76

 savent

-0.75

 תמיד

-0.75

 never

-0.74

 rarement

-0.73

POSITIVE LOGITS

 began

0.76

 suddenly

0.76

 abruptly

0.75

 unexpectedly

0.75

suddenly

0.71

 commenced

0.71

 formally

0.70

 announced

0.70

 unilaterally

0.70

 officially

0.67

Activations Density 0.487%

key information about company or organization names and their related activities

followed by "was"

official announcements and filings

No Comments

No Known Activations