INDEX

Explanations

terms related to summarization and output presentation

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

summaries and summarizing

np_acts-logits-general · gemini-2.0-flash

summary or summarize

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 eletrônico

-0.43

TargetException

-0.41

 interesado

-0.41

ctrons

-0.41

 kteří

-0.40

RunWith

-0.40

 obiek

-0.40

 brugt

-0.40

principalColumn

-0.39

 democrática

-0.39

POSITIVE LOGITS

 summary

1.36

 Summary

1.19

 summaries

1.12

 summarize

1.02

Summary

1.02

 summarizing

1.02

 summarizes

1.01

 summarized

0.99

 overview

0.98

summary

0.96

Activations Density 0.238%

terms related to summarization and output presentation

summaries and summarizing

summary or summarize

No Comments

No Known Activations