INDEX

Explanations

instructions related to thoroughness and carefulness in tasks

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ValueStyle

-0.88

TagMode

-0.85

rungsseite

-0.85

majánló

-0.81

iſchen

-0.81

 témoig

-0.80

<unused8>

-0.79

<unused28>

-0.79

<unused3>

-0.79

[@BOS@]

-0.79

POSITIVE LOGITS

 mucha

0.35

 thoroughly

0.34

 thorough

0.34

 Thorough

0.31

 gerai

0.30

 involved

0.29

 carefully

0.28

 тщательно

0.27

 especially

0.25

Activations Density 0.019%

instructions related to thoroughness and carefulness in tasks

No Comments

No Known Activations