INDEX

Explanations

references to the reader or user in a conversational context

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

Followed by a verb related to action

np_acts-logits-general · gemini-2.0-flash

referring to past actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_16k/average_l0_76

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.31.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

'}),

-0.36

-0.35

an

-0.33

 abgesch

-0.31

vyu

-0.30

こちらも

-0.30

なんと

-0.30

 either

-0.29

 incl

-0.28

mú

-0.28

POSITIVE LOGITS

transQ

0.82

 مرئيه

0.73

styleType

0.70

RegressionTest

0.68

setVerticalGroup

0.68

 betweenstory

0.67

AndEndTag

0.66

новништво

0.65

 createSlice

0.64

makeText

0.63

Activations Density 0.308%

references to the reader or user in a conversational context

Followed by a verb related to action

referring to past actions

No Comments

No Known Activations