INDEX

Explanations

terms related to treatment discontinuation and adverse events

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

stopping or leaving

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

näm

-0.41

NewReader

-0.38

obr

-0.38

 fieltro

-0.36

 lavable

-0.36

 marbre

-0.35

colari

-0.35

visning

-0.34

 diversité

-0.33

jados

-0.33

POSITIVE LOGITS

 Signalez

0.63

 disambiguazione

0.63

writeFieldEnd

0.54

parsedMessage

0.53

fromnode

0.53

 stopped

0.53

 stop

0.52



0.51

TagMode

0.51

transQ

0.51

Activations Density 0.055%

terms related to treatment discontinuation and adverse events

stopping or leaving

No Comments

No Known Activations